Resumé

Beginning with the assumed problem of high turnover among local councilors in Norway, the article compares local political recruitment processes in Finland and Norway. Turnover in Norwegian local councils has proven surprisingly stable at 60-65 percent for a series of elections, whereas it has been significantly lower at 35_—45 percent in the other Nordic countries, including Finland. Turnover among Norwegian councilors is mainly due to voluntary retirement. According to theories of political recruitment and representation, lack of motivation among candidates poses a threat to the democratic quality of political systems, because it undermines the voters' ability to exercise democratic control over politicians. The authors argue that rotation in office need not constitute a democratic problem. On the contrary, empirical evidence is presented to show that participation in political council work may in itself have a politically activating effect on the participants. Very few candidacies can be described in terms of political ambition. Instead, motivation is often created and cultivated through participation. Furthermore, rotation in office may lead to the diffusion of political competence and may therefore constitute an alternative source of democratic control.

Introduction

Democratic government depends on the active involvement of citizens in political matters. Democracy is measured not only by the degree to which actually pursued policy mirrors citizens' opinions, but also by the degree to which citizens actively take part in shaping political decisions. Civic participation will in this article be regarded as a de facto prerequisite for the realization of local democracy. Recently, however, political participation in the Nordic countries has apparently declined. Low voter turnout in the last municipal elections is perceived as particularly worrying in regard to local democracy. Only slightly more than 60 percent of the voters turned out in the last local elections in Finland and Norway, the lowest since World War 11. The Nordic countries have been accustomed to larger turnout at elections than most other European countries.¹ Remarkably high turnover rates among local

---

* Jacob Aars, SEFOS, University of Bergen, Herman Fossgate 5, N-5007 Bergen. Norway; and Audun Offerdal, Department of Administration and Organization Theory. University of Bergen, Christiesgate 17, N-5007 Bergen. Norway.

councilors in Norway are often claimed to be another indicator of declining local democracy.² Turnover is interpreted within a pessimistic view of political recruitment and local democracy. For this reason turnover rates in Norwegian local councils are the point of departure for this article. We want to focus on the comparatively high turnover rates among Norwegian councilors and its possible causes and effects within the framework of recruitment theory and democratic theory.

Recent research has given a more balanced picture of the state of local democracy than the description generally presented through the media. For instance, a recent study by Rose & Skare (1996) indicates that civic participation and support for local government still is high in Norway. Instead it appears that the forms of participation and people's appraisal of local government have changed. Support for local government as a political institution has declined, whereas people now tend to stress the service delivery function of local authorities. This development also signifies a change in people's role orientations towards local government, from the role of citizen to the role of user or consumer. People's participation turns from broad civic involvement in local affairs to greater single issue orientation (Hansen 1995). For local politicians, this development may mean new forms of interaction with local interest groups and organizations. Expanded interrelations between local authorities and groups in the community will most likely present new challenges to local politicians, and recruitment of new candidates might be seen as a key factor in this situation. A continuous supply of willing candidates constitutes a major link between voters and politicians in representative democracies. Increased user or consumer orientation along with a shortage of candidates for office would arguably pose a serious threat to local democracy. High turnover rates among Norwegian local councilors could suggest that this is a plausible interpretation. This article will take a close look at the recruitment of local councilors to evaluate the appropriateness of this description.

Comparison

We wanted to contrast the experiences of Norwegian local councilors with those of councilors in another country in order to shed some light on the nature of local political recruitment in Norway and to assess whether the experiences of Norwegian councilors were particularly negative. Finland was chosen because its municipal structure resembles that of Norway. The number of municipalities is approximately the same - 460 in Finland and 435 in Norway; the average size of municipalities is also similar- 11,000 inhabitants in Finland and 9,000 in Norway. The median population is 5,000 in Finland and 4,400 in Norway. With approximately 12,500 councilors in Finland and

12,700 in Norway, the average number of members in each council is about the same. Both in Norway and Finland there is a council and an executive board, and the relation between them is roughly the same, although members of the executive board in Norway are elected by and among councilors. In Finnish municipalities, executive board members may also be elected among non-members of the council. Another institutional difference is the role of the chief administrative officer. Norwegian CAOs are appointed on the basis of strictly professional qualifications, while Finnish CAOs are elected by councilors on the basis of, among other things, their political affiliation, an insignificant difference in this context.

However, a more significant institutional difference is the electoral system. In both countries, the systems are based on the principle of proportional representation, parties or other groups nominate lists of candidates, party lists dominate the nomination of candidates for local elections, and voters use the list from the preferred party or group as a ballot. The Finnish system is based on personal votes, so voters may vote for single candidates only. In Norway, voters may give an extra vote to a single candidate, omit a name from the ballot, or leave the ballot unchanged in the ballot box. Political parties may also give extra votes to their preferred candidates by listing their names twice on the ballot. This electoral system gives the political parties greater influence over which candidates will eventually be elected. One consequence of this institutional difference is presumably greater public scrutiny of individual candidates in Finland. Candidates in Finnish local elections will to a greater extent than in Norway appear as autonomous political alternatives, while Norwegian candidates to a greater degree will be perceived as party agents.

Political Recruitment

Representative democracies depend on a supply of persons who are willing to work with matters which are defined as being of common concern to the citizens. The attributes of elected persons have been central to studies of political recruitment, especially whether representatives mirror the social background of the voters they represent. Deviations in social background among voters and politicians will be indicative of the openness of a political system, and it will also say something about what norms are being promoted and encouraged through this particular system. Moreover, preoccupation with social background also has to do with perceptions about the quality of the policy being shaped within the political institutions, which turns our attention towards another aspect of the relationship between citizens and representatives, namely mediating and shaping attitudes, and to what degree politicians are substantively representative of the citizens (Peterson 1970).

Since the actual political selection process of candidates has received less attention in recruitment studies, we know less about what criteria are considered important by voters and candidate selectors. Conclusions are in many cases built on comparisons between the social background of voters and their representatives. A problem with these kinds of inferences is that differences in social background, though identified, do not necessarily constitute significant selection criteria for voters or parties. Nevertheless, inferences about the process of recruitment are frequently made on the grounds of data on social or substantive representativeness. Conclusions are also drawn about councilors' attitudes and experiences from certain aspects of the process, for instance the high turnover rate. This implies that turnover is due mainly to local politicians' less agreeable experiences. We want to direct our attention to these kinds of inferences.

Recruitment to elected positions constitutes a crucial link between citizens and politicians in representative democracies. One well-known contribution to a more process-oriented theory of recruitment has come from Prewitt, who describes recruitment of political leadership as a process with several stages (Prewitt 1970, 6-10): Only a share of the population is considered eligible to hold office. Within this group, some establish a dominant social stratum. This group contains another, more exclusive group, the politically active stratum, which is where we find the so-called "recruits," those who aspire to candidacy. Some recruits are nominated as candidates, and the voters eventually decide who will be elected.

Prewitt compares the recruitment process with a Chinese box puzzle: "The Chinese box puzzle suggests an approach to studying the selection of political leaders. There occurs a gradual but continuous process of selection and elimination which narrows the entire population to the few who hold office" (ibid., 7). The Chinese box puzzle metaphor implies that the core box is contained in each of the larger boxes. Consequently, a person who is elected councilor has gone through a process of qualification, meeting the demands at all previous stages in the selection process. Moreover, the continuous screening of candidates involves training and refining the abilities of potential office holders. Arguably, at least two more stages could be added to the selection of political leadership. First, turnover represents another selection process, where some leave office after only one term, and others continue for several terms. Second, within elected assemblies, a selection process will also take place, where some will be appointed for leadership positions while others will remain backbenchers.

The Chinese box puzzle metaphor is not so much a theory as a perspective on recruitment. This perspective does not postulate any relationships between variables assumed to be important to the selection process. Instead, Prewitt's contribution should be seen as a specific frame of reference for the study of political recruitment (Fowler 1993, 56). However, he claims that at each

level, the selection process could be studied both in terms of selection and self-selection. Studies of selection will mainly concentrate on those who select candidates, the demand side of recruitment, while studies of selfselection will mainly focus on features of the individual candidates, the supply side of recruitment.³ The first aspect is basically about how institutions mobilize individual candidates, the other aspect is about how individuals channel their political ambitions.

One stage of crucial importance to the total selection process is the selection of candidates among recruits. To this stage is assigned a particular theory, recruitment theory, which, according to Prewitt, aims to explain "how the politically ambitious focus on particular offices and how political institutions fill the many posts that keep the institutions operating" (Prewitt 1970, 11). Prewitt's perspective can be said to rest on the crucial assumption that there is a relatively large pool of motivated candidates. At least, the number of motivated candidates should be considerably higher than what is needed to fill the candidate lists. It could be argued that personal motivation is of primary importance at this particular point in the selection process. Prewitt's perspective on recruitment thus leans on the notion of political career ambitions. Fowler argues that Prewitt points to a theory of recruitment that is "essentially a theory of careers" (Fowler 1993, 55). Another influential contributor to the field of political leadership selection, Schlesinger (1966; 1994), concentrates explicitly on the role of political ambition, not only for political recruitment, but also for the quality of representative government.

Schlesinger assumes that politicians are driven by the desire for a political career (Schlesinger 1994, 35). The presumption that representatives are driven by their political ambitions is linked to the concept of accountability. If popular control of elected government is to be effective, politicians must care about their political future. Desire to get elected or reelected will work as a disciplinary constraint on representatives, and thereby political ambitions are seen to play an integrative role in representative political systems. Because they want to get elected, politicians will act in accordance with the mandate from the electorate. Turnover is not necessarily a problem within this model. As long as it is a result of the electorate using its opportunity to evict certain representatives, turnover merely confirms that the voters' democratic control is functioning well. But when turnover is mainly due to voluntary retirement, it poses a threat to responsive government. According to Schlesinger, politicians without a desire to get reelected have no strong incentive to act in line with the voters' mandate. Moreover, it is a problem for political institutions if they are incapable of attracting motivated candidates. The competitiveness of elections may seem illusory, and the electorate loses its means to control the representatives.

The theoretical considerations above suggest that recruitment and local
democracy in Norway are under threat. If the recruitment process can be

described as a vicious circle where most politicians leave office after one term and selectors have problems finding willing candidates, local democracy appears to be in danger. Within Prewitt's perspective on recruitment and, more specifically, Schlesinger's theory of political ambition, we will address the problem of turnover among local councilors in Norway. We will focus essentially on possible causes for the high turnover rates, but because retirement is mostly voluntary, we direct special attention to the question of candidate motivation. On the basis of these observations, we shall interpret turnover and discuss the alleged effects of turnover on the functioning of local democracy.⁴

Turnover

Turnover among Norwegian local councilors is high compared to other countries. In spite of some variation among the other Nordic countries, the level of turnover in Norway is persistently higher than in its neighboring countries.⁵ It is by no means unreasonable to assume that high turnover points to a crisis for local democracy. However, high turnover in Norwegian local politics is nothing new. Turnover is registered in the official electoral statistics from 1979 onwards, and the figures display a striking stability at 60-65 percent for recent local elections. A study of six municipalities located on the western coast of Norway shows that the level of turnover has varied throughout elections from 1921 up till today. These temporal data still reveal that the overall level of turnover has been high throughout the mentioned period (Aarethun 1993). Data from a nationwide sample of municipalities in the middle of the 1950s also underline the point that turnover rates have been persistently high among councilors in Norway. These figures show a turnover rate at approximately 60 percent (Aars 1993). If there is a crisis, at least it appears to be permanent.

Although crisis accounts tend to be exaggerated, there still are reasons to believe that the level of turnover could be a problem. If a large number of councilors are replaced at each election, the number of inexperienced councilors will also be high, and the continuity of the policy making process will be affected. The greatest problem of turnover, however, has to do with its causes. If it is mainly brought about by voters and/or parties evicting representatives with whom they are dissatisfied, turnover is not as harmful to democracy as when representatives themselves choose to stand down. If turnover is mainly due to self-selection, the voters' opportunity to exercise democratic control is undermined, and this appears to be the case. One study indicates that the share of turnover caused by eviction, either in the nomination process or in the elections, is less than 10 percent (Aarethun 1993, 105). According to Schlesinger, lack of ambition among politicians will

weaken the links between voters and the elected and lead to unresponsive government. It is therefore a problem for local democracy if local political institutions cannot attract ambitious candidates, or if council work deprives politicians of their initial motivation. Local councilors in Norway mainly leave office voluntarily, and it seems likely that negative experiences constitute the most plausible explanation for the high level of turnover. In other words, councilors are believed to leave politics because they are fed up with it. Since the level of turnover in other Nordic countries is lower, one would have to assume that councilors' experiences there are also more satisfactory. This is the question we want to address through our comparison with Finland. The Norwegian system is seemingly incapable of attracting willing or ambitious candidates. As we have seen, according to the theory of political ambition, responsiveness and accountability might suffer.

The different electoral systems are important to our main assumptions. It has already been proposed that individual candidates will be more visible and their attitudes will be more contested in the Finnish than in the Norwegian electoral system. One demonstration of this difference is Finnish candidates' use of personal newspaper ads to promote their candidacy, a very uncommon campaign strategy in Norway. Single candidates may be promoted as party candidates, or party candidates are presented jointly, but personal ads are hardly ever seen. We assume that these institutional differences will affect the individual candidate's motivation and, consequently, the process of recruitment.

We expect Finnish candidates to be more motivated for candidacy and office than Norwegian candidates, and we assume that the more motivated candidates will find work in office more interesting and rewarding than the less motivated ones. In addition, Finnish candidates aspire for office more than Norwegians. The remainder of this article deals with the reasons candidates give for accepting candidacy and how they vary between the two countries. Are Finnish councilors' experiences more favorable than those of their Norwegian colleagues? Towards the end of the article, we will give an alternative interpretation of the high turnover among Norwegian councilors in particular and of turnover in general.

Motivation

Because of the different electoral systems in Finland and Norway, we expect candidates in Finnish local elections to be more motivated for office than candidates in Norway. The Norwegian vocabulary contains an expression that, directly translated into English, would mean "list fillers."⁶ The expression is commonly known and accepted, and refers to candidates who lend their names to a party or another nominating group. The decision to run by no

means implicates intention of being elected, in faet, they prefer not to get elected. A study including 19 Norwegian municipalities shows that approximately 25 percent of local councilors explain their candidacy in terms of "filling up the candidate list" (Haveraaen & Offerdal 1990). The share of "fillers" is probably considerably higher among candidates than among councilors, and this particular group of candidates probably would not take much interest or pleasure in being politicians. Although they clearly appear to be reluctant candidates, the "fillers" have shown some signs of willingness by eventually accepting candidacy.

Table 1 shows Finnish and Norwegian candidates' reasons for running. The last two last categories were not relevant for Finland, since Finnish candidates cannot be nominated without consent. Nomination for candidacy in Norwegian municipal elections does not require the candidates' consent.

Finnish candidates' reasons for running differ markedly from the reasons given by the Norwegian candidates. A wish to get elected is the main reason among Finnish candidates, given by 74 percent of the candidates. One might expect this to be a necessary qualification for getting nominated at all, but among Norwegian candidates, we make the rather paradoxical observation that more than half of the candidates who accepted nomination had no desire to get elected. If we add those who asked to be let off but were still nominated, 66 percent, i.e., two thirds of the Norwegian candidates, showed little or no desire to get elected. From the point of view of ambition theory, the lack of motivation is alarming.

These findings should be related to the different institutional contexts within which recruitment in the two countries takes place. In the Finnish electoral system, voters vote primarily for individual candidates, secondarily for the party. The opposite is the case in Norway, where voters primarily vote for a party, then, if they wish, they can cast extra personal votes. Accordingly,

individual candidates will be at the center of attention in Finnish local elections, whereas individual candidates in Norway are associated more strongly with their parties. In addition, chances of getting elected once you run are better in Finland than in Norway. On average, 37 percent of the candidates in Finnish municipalities included in this survey were elected. The corresponding figure for Norwegian municipalities is 22 percent.

In both countries, motivation is an important factor in deciding who will get elected. Motivated candidates are elected far more often than those who do not express an interest in getting elected. Still, there is a marked discrepancy between the two countries. Among Norwegian councilors, 34 percent were not motivated, while 13 percent of the Finnish candidates stated no desire to attain office. In other words, one third of Norwegian local councilors never wanted a seat in the local council. In all likelihood, this figure is high compared to any other country.

Experience and Activation

It seems reasonable to suggest a relationship between a councilor's motivation and his desire to continue. On the basis of this assumption, we would expect turnover to be higher among Norwegian than among Finnish local politicians. Nevertheless, this line of reasoning implies a certain view of recruitment which is not necessarily accurate.

As we have noted earlier, a prevailing view of political recruitment presupposes that recruits and candidates actively seek nomination and, subsequently, election. Office is assumed to be attractive and to constitute an incentive for those elected to act in compliance with the voters. Thereby, motivation and ambition - the reasons people get involved in politics - are assumed to precede participation. But few have investigated the opposite relationship, i.e., the possibility that participation creates motivation. Political involvement and ambitions might be thought of as evolving through participation. In the following section we discuss councilors' experiences in office and to what degree these experiences have affected the respondents' political interest and/or level of participation. Following the assumption that voluntary turnover is caused by unsatisfactory experiences in office, we would expect Norwegian representatives to express more negative views of council work.

The persons holding seats in the local councils at the time of the survey
were asked whether council work had stimulated their interest in political
matters.

Among Finnish representatives, 62 percent stated that council work had
greatly stimulated their political interest. This share is smaller among
Norwegian representatives, but judging from the low level of motivation, the

figures presented for Norwegian local politicians nevertheless seem surprisingly uplifting. Although a third of Norwegian councilors showed no motivation to seek office, less than 20 percent stated that council experiences had not stimulated their interest in political matters. Consequently, more than 80 percent reported that their political interest had increased through participation in council work, which does not support a view that council experiences are predominantly negative. Nor does this indicate that all voluntary retirement is due to distress. On the contrary, it appears that many of those who give up council work do so in spite of the gratification. Further inquiry confirms this impression: There is no significant relationship between positive experiences and willingness to run for reelection. The relation is somewhat stronger in Finland (0.16) than in Norway (0.12). Thus, in Norwegian municipalities, 59 percent of those who reported that council work had increased their political interest were still unwilling to run for reelection. The data give little evidence that representatives are unhappy with council work, but councilors resign although they find the work interesting. We shall return to this faet later in the article.

Still it may be conceivable that one's general political interest increases through council experience, but that the desire to continue as a councilor simultaneously diminishes. An interpretation of Table 2 could thus be that participation in council work makes one aware of what political work might have been had it not been for poor working conditions and little visible effect of one's efforts. As a measure of activation, the results from Table 2 should be combined with other measures of actual behavior. If experience as a councilor not only changes one's attitude towards political participation but also affects one's behavior, this suggests that being elected councilor has an activating effect.

One could hypothesize that councilors, through their council experience, have become more active in partisan work or in organizations in addition to having taken more interest in political matters. One powerful justification of local government is its citizen-educative function (Sharpe 1970, 161). Through participation in local affairs, citizens develop valuable political

competence. Experiences in speech, discussion and co-operation further
cultivate faculties that are important for participation in public life, locally as
well as regionally or centrally.

Table 3 shows to what degree we can trace such an activation effect in the Finnish and Norwegian data. The councilors were asked in what ways council work had affected their activities in political parties and organizations: Whether they had been more or less active while in office.

Table 3 clearly demonstrates that council work experience affects the level of activity in parties and organizations considerably. Only a minority of the respondents stated that they had become less active while in office. Participation in council work evidently leads to participation in other political and local affairs. Particularly interesting is the observation that the activation effect is greater among Norwegian than among Finnish councilors. Although Norwegian candidates at the outset are far less motivated than their Finnish counterparts, council experiences seem to enhance participants' interest in politics and encourage further activity outside the councils. From the perspective of recruitment understood as a vicious circle, these are unexpected findings. One might suspect that the mechanism of activation would be stronger in small, tightly knit communities than in large municipalities. Since the Finnish municipalities included in our study are small (see Methodological Appendix), we cannot compare large and small municipalities in Finland. However, comparison between small and large municipalities in the Norwegian sample shows no such relationship.

The relationship between experiences as a councilor and political activation is reinforced by data on those who intend to stand down at the next election. They were asked whether their decision not to run implied withdrawal from all political activity. 49 percent of Finnish candidates and 45 percent of the Norwegians reported they would continue to be politically active. Moreover, it appears that many of the resigning candidates do not give up council work for good, so a large share of the turnover is temporary.

Among Finnish candidates, 22 percent of those with experience from more than one term had been out of office for one term or more before returning, compared to 33 percent in Norway. Thus, although turnover is higher in Norway, a greater share is temporary. This finding also suggests that the lack of experienced politicians following from a high level of turnover gives less reason for concern than one might suspect. Not only do the councilors currently in office possess the necessary experience to fill a seat in the council, many others have acquired such practice from previous involvement.

We have also examined another indicator of whether council work increases political interest and involvement. The degree to which unwilling candidates throughout the election period are being activated into further council service is perhaps the most interesting measure of activation processes. This part of our study was only conducted in Norway because the last Finnish local elections were held in the autumn of 1996, and we could not obtain the Finnish data in time for the subsequent study. In Norway we investigated which candidates (for the 1991 elections) included in our survey sample (conducted in 1994) were also nominated to run in the next elections (in 1995). The candidates were classified into four categories: The Positive, the Activated, the Passive and the Negative. The Positive are those who in 1994 answered that they were willing to run and were actually on a candidate list in 1995. The Activated are those who in 1994 answered that they intended to stand down, but still appeared on the list in 1995. The Passive intended to run, but were not on the list in 1995. The Negative answered in 1994 that they were unwilling to run and were not nominated in 1995. The result of our analysis is shown in Table 4. Councilor and candidate strata are separated, and the table thus distinguishes between candidates who, at the time of the survey, were elected councilors and candidates who were not elected.

Table 4 shows that a major share of the candidates held on to their original
decision about running. The persistently positive or negative represent about
60 percent, but almost 40 percent had changed their minds. Among these, two

thirds have been activated throughout the election period. The result confirms findings from another recent Norwegian study of councilors in municipalities in the county of Sogn og Fjordane (Offerdal & Aarethun 1995), in which activation increased far more than passivity.⁷ Thus, activation was explained in terms of council work experiences. In our current study, we have included the candidate stratum, those who were not elected councilors, as well as the councilors. Quite interestingly, Table 4 shows no difference between the two subgroups. In both groups, a clear majority of those who have changed their minds are being activated. Therefore, activation cannot be explained exclusively in terms of councilors' experiences while in office. It appears that being nominated for a candidate list is enough to activate candidates. For local democracy, this is an optimistic interpretation, because seemingly it takes only very little participation to activate citizens. A supplementary explanation is that many candidates are recruited because of their work on municipal boards and committees, which gets them involved in local political affairs. Anyhow, the evidence supports the interpretation that motivation could be created and nurtured through participation.

The joint data on councilors' behavior rule out the earlier mentioned hypothesis that council participation enhances political interest without generating further actual involvement. Council experience does lead to further participation.

Reasons for not Seeking Election

It has been shown that there is only a weak relation between whether local councilors' experiences while in office are held to be positive and the degree to which they are willing to run again. For this reason, it is hard to maintain that councilors choose to resign primarily because they find council work uninteresting or frustrating. There must be other explanations for the high level of turnover, and we asked the unwilling voters, candidates and councilors why they intended to turn down possible requests to run for local council. The answers are presented in Table 5.⁸

To accommodate comparison, the different reasons have been ranked according to degree of support within each country. Reasons for not seeking nomination will probably vary according to the degree of previous experience with council work. Analysis of different strata in the recruitment process may produce different results as to what reasons respondents give for declining nomination. Table 5 therefore gives information about voters, candidates and councilors separately. The three groups will have different frames of reference for answering the question.⁹ It follows from the logic of Prewitt's Chinese boxes that the voter stratum contains some candidates, and similarly the candidate stratum contains elected councilors.

Four reasons, all related to workload and time commitments, appear to be more important to Norwegian than Finnish respondents. In addition to too much time spent reading documents, council work affects family life, daily work and leisure time. Most of the time-related reasons are also important among Finnish respondents, but three other reasons are clearly more important in Finland than in Norway: First, debates in council follow party divisions too closely. This reason is of particular importance among voters and candidates. Second, seniority reasons are important among councilors. Third, civic duty reasons are equally important to voters, candidates and councilors, and there is a clear relationship between this reason and age in both countries. In Finland, the relationship is particularly strong among the oldest respondents, while the relationship is linear among Norwegian respondents.

We would like to draw attention to the skepticism of political parties which is particularly strong in the Finnish data. The role of the parties constitutes the single most important reason among Finnish candidates and voters for not accepting nomination. These results indicate a conflict between candidates who feel a personal responsibility towards the electorate and the parties. One

possible consequence of the Finnish electoral system is that candidates might easily perceive the electoral mandate as a personal mandate rather than as a party mandate. Recruits drop out because they find it hard to combine the individually oriented nomination and campaigning processes with the partydominated politics of the council.

Some councilors, on the other hand, are apparently able to handle this conflict. If we define seniority reasons by the two first categories in Table 5 ("Have been a councilor long enough" and "Too old"), these reasons are clearly the most important within the councilor stratum. These reasons do not find widespread support among Norwegian respondents. Evidently, many local politicians in Finland decide to step down after having completed a local political career. This is not an important factor for explaining turnover among councilors in Norway. On the contrary, turnover probably explains why seniority reasons are less important among Norwegian councilors and candidates. Where turnover is low, seniority reasons for standing down tend to be more important. Comparing these results with survey results from the other Scandinavian countries shows that Finland resembles Sweden and Denmark in this respect. Seniority reasons are far more important for explaining turnover among Swedish and Danish councilors than among Norwegian councilors (Frandsen et al. 1996, 9ff; Wallin et al. 1981, 412ff).¹⁰ Reasons concerning time commitments are being emphasized in all countries, but nowhere as strongly as in Norway.

It might seem a bit curious that the difference between Finland and Norway regarding time and workload is as evident as it is. It is very unlikely that the explanation can be found in the actual amount of time spent on council affairs. A newly conducted survey in Finland shows that Finnish local councilors on average spend 6.3 hours a week on council business.¹¹ A Norwegian study indicates that Norwegian local politicians might have somewhat heavier time demands, from seven to ten hours weekly (Haveraaen & Offerdal 1990, 97). Nevertheless, this difference is insignificant and can hardly account for the differences concerning time demands between Norwegian and Finnish respondents expressed in Table 5. We also know that the workload of local councilors is an area of dispute in Finland.¹² Studies from Sweden and Denmark do not suggest that the workload of Norwegian local politicians is particularly heavy. It may also be asked whether time spent on council activities is less valued in Norway than in Finland, which is a question of compensation. However, the provisions for economic allowances in the two countries' municipal acts do not indicate important differences at this point. Maybe the perception of time spent on council work is different in Norway, and workload is probably a socially more acceptable reason among Norwegian than Finnish councilors.

It is also worth noticing that reasons for not seeking reelection are not so
much related to a feeling among councilors that council work is not

interesting - or that it is too complicated for ordinary people to carry out. A more likely interpretation of the Finnish answers is that local politicians individually lack the opportunity to influence decisions. Finnish respondents stress that political parties are too dominant, whereas individual representatives are considered to have little influence. Norwegian respondents do not indicate a lack of interest in participation in council or municipal affairs. On the contrary, by stressing reasons related to time, it appears that both recruits and representatives consider the work interesting, but they want to give priority to other activities. An explanation for the differences between Finnish and Norwegian respondents might be that Finnish local politicians, as shown earlier, are more strongly motivated to seek office than their Norwegian counterparts. This interpretation also makes it easier to understand that the reasons given by Finnish respondents are related to the political process itself. Norwegian politicians' reasons for not seeking office relate to lack of motivation. Moreover, Norwegian candidates appear to drift between different societal activities and also between activity and inactivity. Instead of being a career ladder with election for office as the fixed objective, recruitment of councilors in Norwegian municipalities is better described as a process of drift (Barron Crawley & Wood 1991, 43, 53). Furthermore, this description of the recruitment process is not totally compatible with Prewitt's metaphor of the Chinese boxes. "List fillers" obviously skip several stages in the prescribed recruitment process. Instead, it is the actual participation through nomination or through council work that generates further involvement. Even though a major share of councilors quit after a single term, our data indicate that many carry on with political or societal activities in the civic sector. It has also been documented that one out of three resume council work after resigning for one term or more.

Attitudes towards spending time on political work vary between Finland and Norway. It appears that Finnish local politicians, to a greater extent than Norwegians, have chosen to spend time on politics. Thus, they are more strongly oriented towards political careers than Norwegians, and deliberately organize their time with a view to participating in politics. This is also expressed by the fact that councilors, who spend a lot of time on council work, are least likely to state workload as an important reason for resigning.

Recruitment of Local Councilors in Norway - An Interpretation

The starting point for this article was the high level of turnover among local councilors in Norway: How to understand it and how to interpret it. Is turnover indicative of a vicious circle of recruitment in which citizens reluctantly enter local politics, get tired and frustrated, and then leave council

as soon as they get the chance? The circle is completed if we assume that retired councilors share their negative experiences with their fellow citizens. Signs of a vicious circle of recruitment are easily found in data presented in this article. A great many councilors stand down voluntarily, and quite a few of the elected councilors initially had no desire to get elected. According to the theory of political ambition and accountability it is cause for serious concern when candidates lack motivation and are unwilling to seek reelection.

Yet, traces of a benign circle are also visible. Even though the majority of candidates lack motivation to run for office, many of them grow more interested in politics through council work. Quite a few are being activated in the sense that they appear on a list by the time of the next election, contrary to their own claims that they intended to stand down or not seek nomination. These findings confirm results from earlier Norwegian studies. However, the Norwegian data presented in this article have an additional quality compared to earlier studies of councilor recruitment: In addition to councilors, they include candidates. It is demonstrated that the level of activation is the same among not-elected and elected candidates, and it does not take much participation to activate citizens to further political involvement. The results shown in this article are also important for their comparison with Finland. Finnish candidates show much more motivation for office than Norwegian candidates, and a few more Finns also stated that council work had stimulated their political interest. However, the share of candidates that is activated to participate in organizations and political parties is larger in Norway than in Finland. In both countries, about half of the unwilling candidates did not intend to give up political activity despite their decision to stand down.

We have also learnt that neither Finnish nor Norwegian local politicians related their decision to stand down to negative reasons, for instance that they are not interested or that local politics is too complicated for ordinary people. Yet there is a significant difference between the respondents in the two countries. More Finnish respondents related their reasons for not seeking nomination directly to council work, i.e., council debates are too closely tied to party divisions, and the individual councilor has no say. Within the councilor stratum, seniority reasons are important. To Norwegians, the important reasons are all related to time and workload, i.e., too little time for family, work and leisure activities and too much time spent on paperwork before meetings. One might say that Norwegian councilors are oriented towards capacity reasons for standing down, but the reasons for resigning mainly relate to features external to council activities. These constitute the main reasons for voluntary retirements among Norwegian councilors - if we are to believe the reasons they themselves give. There seems to be a conflict of priority between different time-consuming demands that is more strongly felt among Norwegian councilors than among their Finnish colleagues. Thus,

it appears that Finnish councilors deliberately direct more time towards council work. Finns prioritize politics, whereas Norwegians prefer to combine their tasks as councilors with other activities, although council work is interesting.

Offerdal & Aarethun (1995) have earlier used the metaphor conscription, as opposed to career, to characterize recruitment of local councilors in Norway. The conscription metaphor emphasizes the citizen duty aspect of political participation, and stresses another aspect of turnover, namely the fact that turnover in military organizations provides broader links between the population at large and the armed forces. Turnover in political organizations might in a similar manner provide more links between citizens and the relatively few who hold seats in elected assemblies.

The comparison with Finland epitomizes the conscription aspect of local political recruitment in Norway. It can be argued that Finnish municipalities are better perceived in terms of career recruitment as described by Schlesinger (1966). Nevertheless, when we investigated candidates' motives for seeking election, instrumental or policy-oriented motives were important. The opportunity to influence particular issues is the most important motive in both countries. In Norwegian municipalities we find that political interest in general and wish to support one's party are the second and third most important reasons. To Finnish candidates, the opportunity to control the local administration and a chance to direct attention to the group one is representing are considered important reasons for seeking office. These motives relate to different aspects of policy. Finnish candidates are oriented towards the instrumental aspects of politics, they wish to pursue policy goals.¹³ The conscription metaphor receives support in Norway, as many respondents think of it as a citizen duty to engage in local politics when asked. This motive is mentioned by 48 percent of the Norwegian voters, 43 percent of the candidates and 60 percent of the councilors. Corresponding figures in Finland are 32 percent, 36 percent and 39 percent. Discrepancies between the ranked percentages give the same impression. It appears that conscription, or citizen duty, seems to be a suitable way of describing recruitment in Norway, whereas instrumentalism might be a more appropriate description in Finland. Finnish local politicians certainly are more career conscious than local politicians in Norway, but first and foremost their political involvement seems directed towards different aspects of policy, not so much towards gaining office.

It has been mentioned that temporary turnover is higher among Norwegian councilors, and accordingly, Norwegian municipalities appear to be more open to civil society. The relationship between membership in associations and political activity might indicate such openness. In both countries there is a strong relationship, but it is stronger in Norway (Offerdal 1996). People who actively take part in civic activities also participate in local politics, and

recruitment of local councilors takes place in a context of interaction between local associations and local government. Their attention to council work is thus sequential, but one might hypothesize that this is also the case when it comes to other societal activities. Candidates and councilors drift between council seats, positions within local parties and organizational activities. Civil society appears to be highly integrated in local political life, not least through a process of reciprocal turnover.

This interpretation of the recruitment process has possible consequences for future research. It indicates that recruitment studies should employ a wider perspective with regard to the process of candidate selection. Our data seem to suggest a circulation of active persons among several societal activities in the local community. One future research task could be to map the movements among different local government agencies, party positions and leadership positions within local organizations. Another path may be to investigate possible interrelationships between developments within civil society and patterns of local political recruitment. If Putnam's findings of a declining civic life holds true (Putnam 1995), this might also affect recruitment of councilors.

Assessments

If there is anything to our interpretation of councilor recruitment in Norwegian municipalities, it will have bearings on current views concerning the workings of local democracy; both the widely conveyed impression that local democracy is in crisis and the ways to deal with the alleged crisis. It is our contention that the impression of crisis is weakened by the findings presented in this article. Instead, we would like to point out certain characteristics of the Norwegian recruitment process that might be seen as favorable to local democracy.

One effect of turnover is that more citizens gain council experience. Moreover, other personal advantages of political participation, such as political education and development of personal political competence, are more equally shared among citizens if turnover is high. If experiences from council work are generally positive among those who leave office, turnover poses no great threat to democracy.

This kind of interpretation implies another view of accountability than Schlesinger's. The classical elitist doctrine states that accountability must be secured primarily through competitive elections (Schumpeter 1961, 269). Alternatively one might argue that only widespread participation can provide effective representation and responsive government. More widespread participation in local government will lead to more people acquiring the political skills necessary for developing and furthering one's interests.

Phillips (1995) has argued that a politics of presence, i.e., broad participation by many social groups, is more important for popular control of government than the traditional mechanisms of accountability, which she labels a politics of ideas. Rotation in office can thus be seen as a major source of popular control over government from the perspective of a politics of presence.

Our voter survey is in accordance with this reasoning because it shows that more than twice as many Norwegian voters (10 percent) have experience as councilors than Finnish voters (4 percent). This is most likely related to the high turnover rates in Norwegian local councils. It should also be mentioned that the share of voters in the survey who stated that they were nominated as candidates during the last election was 1 percent in Finland and 5 percent in Norway. According to findings presented in this article, even this fairly limited participation has some activating effects.

According to the new Municipal Act in Norway, local authorities are allowed to reduce the number of councilors. The majority of the Municipal Act committee wanted to lower the limit of councilors in each council to 11. Even though the proposal was not carried, the arguments of the committee might be worth noticing. They were among other things:

. . . that in practice there have been certain difficulties in recruiting persons as local councilors or county councilors. In local councils or county councils with many members some of the councilors will usually be passive and have little influence. A reduction in the number of councilors would probably make council work more meaningful and thereby probably have a positive effect on political recruitment at the local and the county level. (NOU 1990:13, 107-8).

A majority on the committee based their reasoning on the supposition that local political recruitment is in crisis. The solution to the defined problem is to have fewer politicians spend more time and energy on local political affairs, i.e., a professionalization of local politics. It appears that the recruitment problem belongs more to the political parties than to local democracy. A comparatively large share of citizens - 16 percent in Finland and 12 percent in Norway - would be willing to run if they were asked. Candidates, in comparison, comprise 1.89 percent of the voters in Finland and 3.30 percent in Norway. The willingness figures suggest that selectors do not have great problems filling the candidate lists. Obviously, it is harder for small than for large parties, since their support is limited. The problem for the political parties is not that the number of willing candidates is too low, but rather how to get hold of those who are willing. The evidence suggests that the parties themselves might be among the greatest obstacles to broader recruitment. In both countries, the parties have nearly monopolized the nomination process. Given this situation, it is interesting to ask to what degree those willing to run are also willing to join a political party. Both in Finnish and Norwegian municipalities, around 25 percent of those willing to run are not willing to join a party, and about 20 percent are not sure, so 45 percent of

those willing to run are either reluctant or unwilling to join a political party.
The figures suggest that political parties and local politics are sometimes
uneasy bedfellows.

In handling the crisis of recruitment, political parties should take into account that political participation does not always require strong commitment. Often the reverse is the case, when political commitment is created through participation. March has argued that often preferences and goals are created or discovered through action (March 1976, 75). For many political recruits, political interest and preparedness for involvement are developed through participation.

Conclusion

Three lessons seem to follow from the evidence presented here:
1) Participation may precede political motivation or ambition. Theories of

recruitment seem to overemphasize the importance of career ambitions. Motivation is often created and cultivated through participation. Among many candidates, accepting nomination is viewed as a citizen duty. Very few candidacies can be properly described in terms of political ambition.

2) Recruitment is not always a process in fixed stages. The metaphor of Chinese box puzzles suggests that one core box, the leadership stratum, is always contained by the outer boxes. Norwegian councilors often appear to skip several stages in the recruitment process. Also, many councilors return from retirement, as implied by the Norwegian concept of "list fillers" and the fact that candidates appear to drift in and out of office, picking up valuable experiences on the way.

3) In Norwegian municipalities, the politics of ideas is widely supplemented by a politics of presence. Through widespread rotation in office, many acquire some degree of political competence, and thus broad recruitment provides a solid foundation for controlling local political elites. Turnover provides relatively broad participation, and widespread participation represents a major potential for holding local political leaders accountable.

---

NOTES

1. Widdicombe Report, table 5.8, p. 146.

---

2. Turnover figures in Norwegian local councils have proven surprisingly stable at 60-65 percent for a series of elections. The level of turnover in the other Nordic countries has been significantly lower, between 35 and 45 percent.

3. Norris & Lovenduski (1995) name these two processes "demand" and "supply." They seem to draw heavily on Prewitt, although he is not quoted.

4. This article focuses mainly on individual candidates. One may also assume that political ambitions are located at the party level. As suggested by Downs (1957, 25), a party's policy may be a function of its desire to gain or retain office. For differences among Norwegian parties in recruitment patterns, see Ringkjøb (1996).

5. It has already been stated that turnover figures in the other Nordic countries are between 35 and 45 percent. In Iceland, turnover after the 1986 local election was 48 percent (Larsen & Offerdal 1994, 82). The level of turnover in the UK and France is about the same, or somewhat lower. In England, a little more than 25 percent of councilors have less than three years of continuous experience (Gyford et al. 1989, 58). In France, one third of the councilors were not reelected, whereas the same goes for about 25 percent of the representatives in the county councils (Mabileau 1996, 71).

6. The Norwegian expression is listefyll.

7. Increased passivity is not necessarily the result of candidates changing their minds. The reason they do not reappear on the list may be that they are left out in the nomination process. In that case, increased passivity may be even less common than table 4 indicates. However, earlier recruitment studies show that very few candidates are turned down in the nomination process. The candidates mainly step down voluntarily (Aarethun 1993; Østrem 1989).

8. Answers from open-ended interviews provide the basis for the categories employed in our survey. The field work was done by Østrem (1989). The first category "Have been a councilor long enough" was omitted from the voter questionnaire.

9. Voters will relate this question to the decision not to seek or accept nomination; to candidates the question will refer to the decision not to seek renomination; to councilors the question will refer to the decision to stand down.

10. See also SOU 1989:108, 67.

11. The survey was conducted by Åbo Akademi. In charge of the project are Krister Ståhlberg and Sari Pikkala.

12. "Time is your greatest problem" was the message from "The Finnish Municipal Reporter" to new councilors after the 1992 election (Finlands kommuntidning no. 1/ 1993).

13. For an elaboration of the instrumental aspects of participation, see Milbrath (1965, 12f).

ACKNOWLEDGMENTS

An earlier version of this article was presented at the workshop on "Local Elites in a
Comparative Perspective," ECPR Joint Sessions, Bern, February 28-March 4, 1997. We wish
to thank the workshop participants for their valuable comments and suggestions.

REFERENCES

Aarethun, T. 1993. Personskifte blant dei folkevalde - ein trussel mot lokaldemokratiet?
Master's Thesis, Bergen: Department of Administration and Organisation Theory.

Aars, J. 1993. Personskifte i lokalpolitikken. Prosjektskisse. Working Paper, Bergen:
Department of Administration and Organisation Theory.

Barron, J., Crawley, G. & Wood, T. 1991. Councilors in Crisis. The Public and Private Worlds
of Local Councilors. London: Macmillan.

Bogason, P. 1996. "The Fragmentation of Local Government in Scandinavia," European
Journal of Political Research 30.

Downs, A. 1957. An Economic Theory of Democracy. New York: Harper & Row.

Fowler, L. L. 1993. Candidates, Congress, and the American Democracy. The University of
Michigan Press.

Frandsen, A. G., Kjær, U. & Pedersen, M. N. 1996. Svarfordelinger fra undersøgelse af
kommunalbestyrelsesmedlemmer der ikke genopstillede ved valget den 16. November 1993.
Kommunalpolitiske studier Nr. 8. Odense: University of Odense.

Gyford, J., Leach, S. & Game, C. 1989. The Changing Politics of Local Government. London:
Unwin Hyman.

Hansen, T. 1995. "Lokalt demokrati ved et vendepunkt?" in Hansen, T. & Offerdal, A., eds.,
Borgere, tjenesteytere og beslutningstakere. Festskrift til Francesco Kjellberg. Oslo:
TANO.

Haveraaen, G, & Offerdal, A. 1990. Opplæring av folkevalde. Bergen: Department of
Administration and Organisation Theory.

Larsen, H. O. & Offerdal, A. 1994. Demokrati og deltakelse. Norske kommuner i nordisk
belysning. Oslo: Kommuneforlaget.

Mabileau, A. 1996. Kommunalpolitik und -Verwaltung in Frankreich. Das "lokale System"
Frankreichs. Basel: Birkhäuser Verlag.

March, J. G. 1976. "The Technology of Foolishness" in March, J. G. & Olsen, J. P., eds.,
Ambiguity and Choice in Organizations. Bergen: Universitetsforlaget.

Milbrath, L. 1965. Political Participation. How and Why Do People Get Involved in Politics?
Chicago: Rand McNally & Company.

Norris, P. & Lovenduski, J. 1995. Political Recruitment. Gender, Race and Class in the British
Parliament. Cambridge University Press.

NOU 1990:13. 1990. Forslag til ny lov om kommuner og fylkeskommuner. Oslo:
Forvaltningstjenestene.

Offerdal, A. 1996. Organisasjonar og politikk. Working Paper, Bergen: Department of
Administration and Organisation Theory.

Offerdal, A. & Aarethun, T. 1995. "Rekruttering av folkevalde: Plikt, engasjement og
karriere," in Hansen, T. & Offerdal, A., eds., Borgere, tjenesteytere og beslutningstakere.
Festskrift til Francesco Kjellberg. Oslo: TANO.

Peterson, P. E. 1970. "Forms of Representation: Participation of the Poor in the Community
Action Program," American Political Science Review 64:2.

Phillips, A. 1995. The Politics of Presence. Oxford: Clarendon Press.

Prewitt, K. 1970. The Recruitment of Political Leaders: A Study of Citizen-Politicians.
Indianapolis, New York: The Bobbs-Merrill Company, Inc.

Putnam, R. D. 1995. "Bowling Alone: America's Declining Social Capital," Journal of
Democracy 6:1.

Rhodes, R. A. W. 1996. "The New Governance: Governing Without Government," Political
Studies.

Ringkjøb, H. E. 1996. Rekruttering til lokalpolitikken. Ei undersøking av partia sitt
rekrutterings- og nominasjonsarbeid i fire norske kommunar. Masters Thesis, Bergen:
Department of Administration and Organisation Theory.

Rose, L. & Skare, A. 1996. Lokalt folkestyre i klemme? Noen momenter vedrørende
innbyggernes forhold til kommunen og kommunepolitikken. Paper presented at conference
on local democracy, Trondheim, August.

Schlesinger, J. A. 1966. Ambition and Politics. Political Careers in the United States. Chicago:
Rand McNally & Company.

Schlesinger, J. A. 1994. Political Parties and the Winning of Office. Ann Arbor: The
University of Michigan Press.

Schumpeter, J. A. 1961. Capitalism, Socialism and Democracy. London: George Allen &
Unwin Ltd.

Sharpe, L. J. 1970. "Theories and Values of Local Government," in Political Studies XVIII:2.

Wallin, G., Back, H. & Tabor, M. 1981. Kommunalpolitikerne!. Rekrytcring - arbetsforhållanden
- funktioner. Del 2. Rapport 8 från kommunaldemokratiska forskningsgruppen.
Kommunaldemokratiska kommittén, Ds Kn 1981:18. Stockholm: Kommundepartementet.

Widdicombe Report. 1986. The Report of the Committee of Inquiry into the Conduct of Local
Authority Business: Research Volume IV: Aspects of Local Democracy. (Chairman Mr.
David Widdicombe QC) Cmnd 9801, London: HMSO.

Østrem, W. 1989. Personutskiftninger og arbejdsvilkår i kommunalpolitikken. Masters Thesis.
Bergen: Department of Administration and Organisation Theory.