The Scandinavian political systems have traditionally been characterized by large mass parties representing the basic social cleavages in these societies. Thus, in 1947, 27 percent of the Danish electorate were paying members of a political party. By 1988, however, the ratio had dropped to 7 percent (Elklit 1991).

This dramatic decline, which naturally is a matter of great concern among party leaders, also raises a series of questions of a theoretical nature. These questions concern first and foremost the functions of political parties in the Danish political system. In the existing literature on political parties a variety of party functions have been identified, the most important being the political mobilization and socialization of the mass public, articulation and aggregation of interests, recruitment of political leaders and formation of public policy (cf. Macridis 1967; King 1969; von Beyme 1985; Pedersen 1989). The large class parties distinguish themselves from other parties by the importance of the mobilizing and socializing functions, making these functions the natural focus of an analysis of declining party membership. If Danish political parties, as some observers argue, are changing from mass parties to media parties (Pedersen 1989), will the parties still be able to perform their traditional mobilizing and socializing functions? If not, have other structures or modes of participation replaced the parties? And

Resumé

Declining party membership in Denmark is analysed in light of the general development of political participation in the 1970s and 1980s. It is demonstrated that the decline in party membership had nothing to do with a general decline in participation. The decline is rather the result of three different processes: (1) the declining number of farmers, (2) the weakening of the organization of the workers, and (3) the political mobilization of the new middle class and women. It is argued, therefore, that the causes of the decline are primarily demographic and socio-economic. It is furthermore argued that the declining membership threatens the traditional mobilizing and socializing functions of the parties and thereby may increase political inequality in the Danish society.

if so, how have these affected the hitherto high level of political equality
in Denmark?

The aim of the article is to describe the precise nature of declining party membership in Denmark and to use this as the basis for a discussion of the causes and consequences of the decline. The strategy of analysis will be to look at the changes in formal party membership in relationship to other changes in political partcipation in Denmark over the last two decades.

A central question is whether the decline in party membership is a manifestation of a general trend towards lower political participation or a manifestation of a change in the way people participate in politics. Theories about the role played by the media lead us to expect a decrease in political participation in general. There are, however, also good reasons to believe that declining party membership goes hand-in-hand with greater participation other areas of political life. For example, the level of education in Denmark has increased considerably during the last 20 years. In 1971 only 4 percent of the population had obtained a high school diploma, compared with 20 percent in 1987.' Some scholars have argued that better educated people do not need the parties as socializing agents or as a help to orient themselves to political questions. These scholars point to the cognitive mobilization of people with a higher education (Dalton 1984; Inglehart 1989; Poguntke 1990). Cognitively mobilized groups will manifest high political activity, but mainly through organizations less bureaucratic than political parties.

Another question is whether there is a trend toward increasing social inequality in political participation. If declining party membership is first and foremost a manifestation of a change in the modes of participation selected, with more 'up-to-date' modes replacing more traditional ones, we might well expect the change to be accompanied by growing social differentiation. Compared to most other countries in the world, the Nordic countries have until now had a high degree of social equality in political participation, primarily as a result of the mobilizing activities of large class organizations for farmers and workers (Rokkan & Campbell 1960). If these organizations are weakened, it is reasonable to expect increasing social inequality in participation.

Yet another central question, closely connected to the former, is whether a trend toward increasing social differentiation with respect to modes of political participation exists. Do people participating in the new modes differ from people participating in more traditional modes? The literature on cognitive mobilization leads us to expect that the growing number of highly educated people will use modes of participation other than those used by farmers and workers.

These three questions guide the empirical analyses reported upon here.
The first section deals with the overall development of the different modes

of participation and with the relationship between them over time. The next two sections deal with different aspects of social differentiation relating to education, social class and gender. On the basis of these analyses the final section discusses the causes and consequences of declining party membership with particular emphasis placed on the mobilizing and socializingfunction parties.

But before turning to the analysis, some comments about the data employed are warranted. Unfortunately, relevant data prior to 1971 do not exist. The national election of that year was the occasion for the first comprehensive political survey conducted in Denmark (Borre et al. 1976). Thus it is not possible to reach back in time to the decade when the biggest decline in party membership occurred. This data set, furthermore, is less than complete with respect to our needs, lacking information on such important factors as political participation through union and grass-roots activities. The 1971 data nevertheless provide a baseline, however incomplete, which to trace the subsequent development as revealed by more complete data sets from two later points in time - 1979 and 1987 respectively.²

The Extent of Political Participation

It is an open question as to whether or not simple membership in a political party should be considered as political participation. For many party members, participation is limited solely to paying the membership fee. Even so, it seems reasonable to argue that the mere act of paying a membership fee is evidence of some sort of political involvement. In any event, substantial attention has been devoted to membership figures. It is therefore appropriate to compare the development of party membership over time to the development of participation in unions and in grass-roots activities - the other two most important modes of political participation in the Nordic countries.

Fortunately, complete membership figures are available, based upon information from the political parties themselves, dating back to the end of the 19405. Fig. 1 shows that formal party membership in Denmark has decreased dramatically since the Second World War, the sharpest decline taking place in the 19605. The figure also shows that the two large class parties in Denmark - the Liberals and the Social Democrats - have suffered the greatest decline in party members. None of the four old parties has escaped the general trend, however.

Data from surveys, presented in Table 1, are in keeping with the picture
described here. These data also reveal a sharp decline in active membership
as measured by participation in party meetings, often considered the very

essence of the Danish democratic system. While there is hardly any doubt that the decline in political participation in terms of party membership has been amplified by a corresponding decline in active party membership, the passive/active ratio should, however, be interpreted with caution because the wording was not identical in the three surveys.³

The development has, in fact, been even more pronounced than indicated by these figures because, as shown in Table 2, the decline in party membership primarily a result of strong generational changes, with the younger generations exhibiting very low membership figures (about 5 percent). Contrary to the life-cycle hypothesis argued by Kristensen (1980), people do not become more prone to holding formal membership of a political party when they grow older. The young people of today are, and will presumably remain, at the low end of this mode of political participation,

Union activity also appears to have declined from 1979 to 1987 despite a slight membership rise (cf. Table 1). However, union activity is probably exposed to stronger short-term changes over time than other modes of participation, making it difficult to identify more general tendencies on the basis of only two measurements.

Strictly speaking, grass-roots activities are not new modes of political

participation in Denmark, since they appeared as early as the last century in connection with the large-scale political mobilizations of social groups. Still, when grass-roots activities emerged in Denmark in the 19705, they were seen by many observers as new and different phenomena. It may be argued that they reappeared in the 1970s because of their strong mobilizing and socializing potentialities (Svensson & Togeby 1991). Grass-roots participationisdefined

ticipationisdefinedhere as participation in collective political activities open to everybody and thus not restricted to members of particular groups or organizations. The concept is operationalized by four forms of activity: (a) participation in public meetings, (b) participation in demonstrations, (c) giving economic support and (d) signing a petition. Starting from a very low level at the end of the 19605, grass-roots participation has increased gradually throughout the 1970s and the 1980s to a fairly high level in 1987.⁴

This leads to the conclusion that there has been no general decline in political participation in Denmark in recent years. Participation through the established and conventional channels has declined, but this has been compensated for by a corresponding increase in less conventional political activities. This conclusion is supported by the fact that voting turnout has not diminished and that interest in politics, as measured by consumption of political news, has remained stable at a reasonably high level (cf. Table 1).

When talking about the interwar period and the early post-Second World War years as the era of the classic party system, this not only suggests that the four old parties (Social Democrats, Social Liberals, Liberals and Conservatives) won most of the votes and thus constituted the basis for government, it also suggests that the four old parties were the decisive unifying and mobilizing forces in the Danish population. This is illustrated not only by high membership figures, but also by the large number of other activities connected with party membership (Pedersen 1989). The declining party activities indicate that the parties now have problems with their mobilizing and organizing role, and that this role may well have been taken over by less formalized grass-roots activities. This development raises a number of questions about the social distribution of involvement in different modes of participation.

Participation and Social Resources

Differences in political participation have traditionally been explained by reference to differences in political resources. A comparative perspective, however, illuminates large differences in the level of social equality between countries. Verba et al. (1978) proposed a solution to this puzzle by making a distinction between individual and collective resources. Individual resources are, for instance, money, education, status, social relations and time. Such resources are connected with psychological factors like political interest and political efficacy. Collective resources, on the other hand, are membership in organizations and identification with a class or a social group. Collective resources may compensate for a lack of individual resources.

The theory is well illustrated by the famous study by Rokkan & Campbell (1960) which showed striking differences between Norway and the United States regarding the social distribution of political participation - with the highest degree of equality in Norway. These differences are attributed precisely to the fact that Norway has strong class-based organizations for farmers and workers, while the US does not.

The situation in Sweden and Denmark has been quite similar to that in Norway. Today, however, the political parties in Denmark are severely v/eakened. If the theory about the importance of collective resources for political equality is valid, we should expect that this weakening of the parties has caused a growing inequality in all modes of political participation in Denmark. If the theory about the cognitive mobilization of the welleducated of the population is valid, we should also expect increasing social differentiation with regard to the modes of participation.

Education is obviously an important individual resource for political participation. A high education creates skills and competence which can be used for political activity and also develops the motivation and interest leading to the use of these skills for political activity. Based solely on the individual resources a positive relationship between education and participation is therefore to be expected. In Table 3, however, it is shown that the relationship as regards party membership is quite different, pointing to the importance of collective resources. The lowest level of participation is found among people with high education resulting in a reverse kind of inequality. By 1987 the difference had narrowed, but not sufficiently to change the basic pattern.

Unions are the creation of the working class, so a reverse inequality should also be expected here. However, in 1979 all groups were roughly equally active, whereas in 1987 the well-educated were most active. This implies a slight tendency towards greater inequality, in favour of the welleducated.

Regarding grass-roots participation, on the other hand, we see a tendency towards greater equality between 1979 and 1987, although the level of inequality remains high. In fact, it is much higher in this area of participation than in any of the other areas. It is unlikely, moreover, that these differences are due solely to differences in resources. Grass-roots activities were important in mobilizing students and the new middle class in Denmark in the 19705, and the strong inequality in 1979 is undoubtedly a result of this uneven recruitment. The developments in the 1980s, by comparison, demonstrate a slight weakening of this wave of mobilization.

Finally, with respect to reading about politics in the newspapers, we find a pattern of social inequality very similar to inequality in grass-roots participation. Altogether then, Table 3 indicates a modest tendency towards greater equality in connection with grass-roots participation and newspaper

reading, and an equally modest tendency towards greater inequality in
connection with activities in the parties and the unions.

Some characteristics of these developments become more apparent when the relationship between occupation and political participation is examined. In Table 4 the respondents are first divided into three categories - i.e. workers, white-collar employees and self-employed. These categories are then subdivided according to education and sector of employment. Conceptually 'new middle class' is defined as highly educated persons employed within education, health care and other public service institutions, is here operationalized as public sector employed white-collar workers with at least a high school diploma.

If we focus first on party membership, we see that the figures for selfemployedpersonsin
i.e. farmers, were very high in 1971 (54

percent) and have remained surprisingly high since then. Therefore, the problems facing the Liberals must be attributed to the declining number of farmers and not to a declining ability to organize the farmers. Whereas 27 percent of the labour-force was employed in agriculture in 1950, the corresponding figures were only 11 percent in 1970 and 6 percent in 1987

(Goul Andersen 1984; Statistisk tiårsoversigt 1989). This conclusion fits well
with the observation that the electoral behaviour of the Danish farmers has
for many years been extremely stable (Glans 1989).

Party membership figures for workers in 1971, by comparison, were much lower: 18 percent among skilled workers and 15 percent among unskilled workers. These figures can be compared with an overall membership for the Social Democrats in 1950 of 35 percent (Elklit 1991). But as late as 1971 the level of organization of the workers was much higher than the level of organization of the white-collar employees. Since then the organization percentage for the workers has declined by almost one-half.

Turning our attention to the new middle class, we see that party membership and always has been, at a modest level. Taking into account the fact that the new middle class in general is very active in politics, it is conspicuous that their activities do not include membership in political parties. It is well documented that the 1970s was a period with strong political mobilization of the new middle class, a mobilization which continued the 1980s, although at a slower pace (Svensson & Togeby 1991). But this mobilization has largely taken place without involving the political parties.

Trade unions were originally an integrated part of a united labour movement, and for many years the members were predominantly bluecollar Organization of white-collar workers came later, but today they are as well organized as the blue-collar workers. As we see from Table 4, the level of union activity in 1979 was quite high among both skilled workers and public sector white-collar employees. In the course of the 1980s, however, skilled workers became less active, whereas public sector employees remain as active as in the 19705. During recent years, in short, public sector employees seem to have replaced the workers as the vanguard of the trade unions.

The greatest social differences are nevertheless to be found within grassroots As already mentioned, grass-roots participation is linked to the political mobilization of the new middle class in the 19705. In the 1980s mobilization has weakened, and grass-roots participation has become more generally accepted, but the social differences are still impressive.

Developments over the last two decades can be seen in two different perspectives. On the one hand, we can emphasize an aggregate picture based upon all modes of participation. On the other hand, we can look at them one by one. In this latter perspective it appears that all things considered, the changes have been relatively moderate. Social inequality in political participation has not increased very much over the last 20 years.

This conclusion is contrary to the expectations suggested by the theories

stressing the importance of collective resources for political participation. The weakening of political parties should have had long-term consequences, manifest in growing inequality in all modes of participation, but this is not so. These surprising results might be explained by an increase in individual resources experienced by social groups usually considered to lack resources. A skilled worker today will typically have spent 10 years in primary and secondary school, attended technical school for a considerable period and, in addition, participated in a number of more specialized courses. If formal education is an important political resource, the political resources of the skilled labour-force have increased considerably in the last 20 years.

The conclusion to be drawn, however, is totally different if we look at all forms of political activity and the total amount of participation. The tendency in the 19605, the 1970s and the 1980s appears to have been to replace party activities with grass-roots activities. This indicates that a mode of participation characterized by social equality, or even by a reverse inequality, has been replaced by a mode of participation characterized by a distinctive social inequality favouring those with higher education. The more predominant grass-roots participation is in terms of all forms of participation, the stronger the overall social inequality in political participation. is here that we see the dramatic consequences of the crumbling of the collective resources of workers and farmers. In Sweden, where the traditional organizations are much stronger than in Denmark, we find that social inequality In grass-roots participation is correspondingly lower (Togeby 1989).

To this can be added that the number of persons with a high school diploma and a long theoretical education has increased a great deal since the 19605. If 2 percent of the party members in 1971 had a high school diploma, 42 percent of the grass-roots actives have one in 1987. The increase in the level of education has by itself caused an increase in social inequality.

Despite modest changes in the individual modes of participation throughout 1970s and 1980s, in short, it would seem that political participation in Denmark has on the whole moved towards greater social inequality. The overall level of participation has not declined, but the participants have changed. Inasmuch as loosely organized grass-roots activities seem to have taken over most of the mobilizing and socializing functions in the Danish political system, moreover, the primary beneficiaries appear to be the new middle class and other well-educated people.

Participation and Gender

At the aggregate level of analysis, one of the greatest gaps with respect to
participation in political life has been that between men and women

(Lafferty 1980). This phenomenon can be attributed to women's relatively modest possession of political resources, related to female upbringing and women's position in society. In Denmark men have, until quite recently, also participated much more than women, but today these differences have almost vanished at the mass level. Strong differences are still found at the elite level, but not at the mass level, and the changes have, as indicated in Table 5, largely taken place during the 1970s and 1980s.

The general development has been identical for all modes of participation. difference between male and female participation has diminished all cases. As of 1987, the largest difference to be observed appears in connection with people's own evaluation of how often they read about politics in the newspapers. This corresponds to the fact that more men than women claim to be very interested in politics. Party activity also continues to be characterized by gender differences, with men dominating among the active party members. Women only constitute one-third of the members frequently attending the party meetings. On the other hand, we find an equal amount of participation among men and women regarding union and grass-roots political activities.

Overall the gender gap in political participation has now been reduced

to the point where it is no longer an issue at the mass level. This applies to all modes of participation. At the same time a mode of participation characterized by great inequality (party membership) has been replaced by a mode characterized by total equality (grass-roots participation). In contrastto conclusion reached about the growing social inequality in participation, the conclusion on gender inequality is that it has diminished and almost disappeared.

In Table 5 it is also shown that decreasing differences between men and women are in general due to an increase in female participation, the only exception being party activity. Political parties aside, participation of women has increased in absolute as well as in relative terms. The 1970s and 1980s, in other words, witnessed a considerable political mobilization of women, but this mobilization was accomplished without the assistance of the parties.

The political mobilization of women can in large part be traced back to their growing integration in the labour-market (cf. Togeby 1991). In 1960 only 44 percent of Danish women between 15 and 64 years of age were wage-earners. In 1970 this proportion had increased to 60 percent, in 1979 to 70 percent and in 1986 to 77 percent (Historical Statistics 1960-84, 1986; Historical Statistics 1960-86,1988). Today almost all women in the younger generation are wage- or salary-earners, and very few women leave the labour-market when they have children. Work outside the home has become a lifetime responsibility for women, just as it is for men.

Why, one may ask, was the mobilization of women in the 1970s and 1980s not of benefit for the political parties? The explanation seems to be twofold. First, the mobilization of women was part of a major mobilization wave in the 1970s involving first and foremost the new middle class. These women were well-educated and therefore equipped for cognitive mobilization. Assistance from the political parties was not needed. Second, the parties missed their opportunity at the time when women entered the political arena. Traditional party organizations were the essence of patriarchal structures disliked by young and politically engaged women. In addition, the costs were smaller and the benefits greater for women who involved themselves in the local union or in a grass-roots organization. And that is what they did in the 1970s and continued to do in the 1980s, even although the parties eventually recognized the need to pay attention to the women.

The Causes and Consequences of Declining Party Membership

The analysis up to this point can be summarized by suggesting that declining

party membership is the result of at least three different processes. First, there has been a radical decline in the number of people working in agriculture. The declining number of people in agriculture, not the declining level of organization among individuals is to blame for the erosion of one of the old pillars in the Danish party system. Second, the level of organization among workers has dropped considerably, although the greatest change goes back to the period prior to the one analysed here. Third, the large mobilization waves of the 1970s did not benefit the political parties. The new middle class, the well-educated and women became very active in politics, but they manifested little interest in the parties. These cognitively mobilized people were attracted, instead, by loosely organized grass-roots activities.

This leads to the conclusion that declining party membership in Denmark must primarily be attributed to demographic and social class factors and to the rising level of education. Various other factors mentioned in the relevant literature, such as changes in the media structure, the municipal reform of 1970, high membership fees, and the small number of elected offices (Sainsbury 1983; Pedersen 1987, 1989; Sundberg 1987, 1989; Elklit 1991), have - at most - only accelerated or reinforced a development in progress.

It is possible that the changing media structure has had some effect on party membership, but if so, only on the declining membership of workers. The argument that television keeps members away from party meetings because political information is communicated with greater skill and higher entertainment value by the networks than by the local politicians, would imply a general decline in political activity, but this is not so. We have to go a long way back in history to find a period where so many people attended political meetings and demonstrations as in the 1970s and the 1980s. There is no evidence whatsoever that people prefer an armchair in front of their television to political activities undertaken together with other people. Only the political parties are left out. Most people still find it more interesting to be part of a social event than to watch it on a screen.

Neither did the municipal reform in 1970 have a particularly negative impact. The reduction in the number of administrative units had the most dramatic consequences in thinly populated rural areas. We should therefore expect the farmers to be the most affected by the municipal reform, but there has been no decline in their membership ratio since the reform. In the case of the workers, the largest decline occurred before 1970.

The same kind of objections can be raised against the theory, formulated by Sundberg (1987, 1989), which attributes declining party membership to the lack of elective offices available for Danish party members. If this problem arose in connection with the municipal reform in 1970, it certainly did not cause any important changes with regard to party participation. If, on the other hand, the lack of elected offices has always been a problem

in the Danish party system, this lack cannot explain the decline in party membership. This is not to deny, however, that the large number of elective offices in Finland and Norway may have helped to counteract a development similar to that seen in Denmark.

The only phenomena demanding a special explanation are the declining organization of the workers and the weakened party organization of the Social Democrats. Also, in this case it is possible to point to a number of social structural factors. An important factor in this regard may be found in the changing housing pattern. The old working-class neighbourhoods have almost disappeared and manual and non-manual workers live sideby-side the suburbs. Their children frequent the same schools and play football in the same clubs. Consequently, the basis for a distinctive workingclass has crumbled. The general increase in schooling and the possibility of educational mobility also pull in the same direction. An additional explanation is presented by Sainsbury (1983) pointing to the fact that the Swedish Social Democratic party has consciously worked to strengthen its organization, whereas the Danish party realized the seriousness the situation much too late.

This leads us to the question about the consequences of declining membership political parties in particular and, by implication, for the political system in general. Do the parties have any reasons - other than economic - to be concerned about the loss of party members? Is the loss critical to the functioning of the political parties?

Most scholars agree that political parties are changing from being mass and class parties to being media parties, but they disagree on the appraisal of the development. Pedersen has provocatively asked: 'What is a party to do with so many members? Are they at all necessary in times when the party activities are increasingly financed by means of taxes, and when parties no longer need activists for practical tasks? Is a party organization not an impediment in the era of minority government?' (1989, 276). Yet a phenomenon that is not even mentioned by Pedersen is the traditional mobilizing and socializing functions of the parties. As the analyses presented this article have demonstrated, at least these functions have been seriously damaged during recent decades.

In Table 1 it is shown that the decline in party membership was followed by a decline in the number of active party members, i.e. members frequently attending party meetings. It is precisely the active party members that are the agents of mobilization and socialization. Table 6 illustrates that active party members both in 1971 and in 1987 differ decisively from the ordinary voters by a higher level of political involvement. Members are more likely to claim an interest in politics, to feel politically competent, to persuade other people to vote for a certain party, and to be less influenced by election campaigns. Consequently it seems safe to conclude that declining party

membership has resulted in fewer core members, fewer agitators and fewer
stable voters.

Specifically, this seems to be a problem for workers and for the Social Democratic party. Workers have traditionally been mobilized by the political and workers are not generally attracted by grass-roots activities. might be correct, as argued by Elklit (1991), that there is no direct relationship between the number of party members and the number of votes in general elections. However, some kind of relationship is nevertheless indicated if we look at the long-term development of the Social Democratic party in Denmark. Glans (1989) has shown that whereas 73 percent of all Danish workers in 1953 voted for the Social Democrats, only 41 percent did so in 1988. It is understandable, therefore, that the Social Democrats worry about the declining level of membership. It might be possible for a party to receive financial support and to appear on the front pages of the newspapers without a large number of members, but it is not possible to remain the big, stable and dominating centre in Danish politics if the party has to rely solely on a small professional, media-oriented party organization.

We can also note in passing that Table 6 illustrates a growing gap between the ordinary citizen and party members, making it increasingly difficult to perceive party members as representatives of ordinary voters. If the party organizations are run by political specialists, for whom politics is a profession, it will, as Pedersen (1989) also has argued, be difficult for the party organizations to invoke a higher democratic legitimacy than any other self-appointed group in Danish society. Presumably this will add even more to the decline of the political parties.

Another point emerging from the analysis is a growing social inequality in political participation which is at least in part attributable to declining party membership. The well-educated have taken a much more prominent position among the politically active than earlier. Nobody knows how much effect the social composition of the participants has on the content of public decisions, but the few studies on the subject which do exist suggest that this is not an unimportant factor (cf. Verba & Nie 1972; Verba et al. 1978; Klingemann 1985; Kaase 1990).

The large number of well-educated people loudly articulating a wide variety of demands clearly influences the political agenda. These activists are, of course, assisted by the media, but the initiative rarely comes from the media themselves (Kingdon 1984). Pedersen is possibly right when claiming that the parties have no other alternatives than to become media parties, managed by a market-orientated sales organization. But it is not true that 'the little number of party leaders interacting with the professionals of the media world set the political agenda' (Pedersen 1989, 277). There will still be a great many demands from what has been labelled 'the grass

roots', and politicians will be forced to relate to these demands. However,
these demands may rise disproportionately from the interests of the welleducatedand
shaped by this group's perception of the world.

When looking at the decline of the political parties in the perspective of the general development in political participation, the consequences for the political system seem rather serious. It would be interesting, therefore, to be able to offer some good advice to the parties about how to restore strong party organizations, but the era of the large mass parties and class parties in Denmark has probably gone forever. The last opportunity to revitalize the parties was perhaps presented in the 1970s by the political mobilization of the new middle class and of women, but the parties failed to take advantage of that opportunity.

---

NOTES

1. These figures are drawn from national election studies.

2. The 1979 data are primarily from the so-called 'Mass survey' (Damgaard et al. 1980), supplemented by data from the election study (Borre et al. 1983). The 1987 data are constructed by adding the data from the election study (Elklit & Tonsgaard (eds.) 1989) to the data from a study of grass-roots participation in the Nordic Countries (Togeby 1989).

3. Unfortunately, the wording of almost all the questions about participation has changed from survey to survey. This creates some problems when comparing levels of participation. concerning the relationship between participation and other factors are undoubtedly more reliable.

4. In both the 1979 and 1987 surveys grass-roots participation was defined by the same four activities, although there were some differences in the wording (ef. Svensson & Togeby 1989). Yet other studies confirm an increase in grass-roots participation in the 1980s (Svensson & Togeby 1991). Unfortunately no data exist on grass-roots participation in 1971. One of the reasons for measuring it for the first time in 1979, in fact, was precisely that only during the preceding decade did grass-roots activities rise to a level where they could no longer be ignored.

REFERENCES

Beyme, K. von 1985. Political Parties in Western Democracies. Aldershot: Gower.

Borre, O. et al. 1976. Vælgere i 70'erne. Resultater fra interviewundersøgelser ved folketingsvalgene
1971, 1973 og 1975. København: Akademisk forlag.

Borre, O. et al. 1983. Efter vælgerskredet. Analyser af folketingsvalget 1979. Århus: Politica.

Dalton, R. J. 1984. 'Cognitive Mobilization and Partisan Dealignment in Advanced Industrial
Democracies', Journal of Politics 46, 264-284.

Damgaard, E., ed. 1980. Folkets veje i dansk politik. København: Schultz.

Elklit, J. 1991. 'Faldet i medlemstal i danske politiske partier. Nogle mulige årsager', Politica
23, 60-83.

Elklit, J. & Tonsgaard, 0., eds. 1989. To folketingsvalg. Vælgerholdninger og vælgeradfærd
i 1987 og 1988. Århus: Politica.

Glans, I. 1989. 'Langtidsudviklingen i dansk vælgeradfærd', in Elklit, J. & Tonsgaard, J.,
eds., To Folketingsvalg. Vælgerholdninger og vælgeradfærd i 19870g 1988. Århus: Politica.

Goul Andersen, J. 1984. 'Udviklingen i sociale modsætningsforhold frem mod år 2000', in
Konflikt og tilpasning, Egmont Fondens Fremtidsstudie. København: Aschehoug Dansk
Forlag.

Historical Statistics 1960-84, 1986. Paris: OECD.

Inglehart, R. 1989. Culture Shift in Advanced Industrial Society. Princeton: Princeton University

Kaase, M. 1990. 'Mass Participation', in Jennings, M. K. & Deth, J. W. van et al., eds.,
Continuities in Political Action. New York: de Gruyter.

King, A. 1969. 'Political Parties in Western Democracies: Some Sceptical Reflections', Polity
2, 111-141.

Kingdon, J. W. 1984. Agendas, Alternatives and Public Policies. Boston: Little, Brown.

Klingemann, H.-D. 1985. Formen, Bestimmungsgrunde und Konsequenzen politischer Beteiligung.
Forschungsbericht. Zentralinstitut fur sozialwissen-schaftliche Forschung der
freien Universität Berlin.

Kristensen, O. P. 1980. 'Deltagelse i partipolitiske aktiviteter', in Damgaard, ed., Folkets
veje i dansk politik. København: Schultz.

Lafferty, W. M. 1980. 'Sex and Political Participation: An Exploratory Analysis of the 'Female
Culture', European Journal of Political Research 8, 323-347.

Macridis, R. C. 1967. 'Introduction: The History, Functions, and Typology of Parties', in
Macridis, R. C, ed., Political Parties. Contemporary Trends and Ideas. New York: Harper
& Row.

Pedersen, M. N. 1987. 'The Danish "Working Multiparty System": Breakdown or Adaption?',
in Daalder, H., ed., Party Systems in Denmark, Austria, Switzerland, The Netherlands and
Belgium. London: Frances Pinter.

Pedersen, M. N. 1989. 'En kortfattet oversigt over det danske partisystems udvikling", Politico
21, 265-287.

Poguntke, T. 1990. Unconventional Participation in Party Politics. The German Experience.
Paper, ESF/ESRC Conference on Political Participation, Manchester.

Rokkan, S. & Campbell, A. 1960. 'Citizen Participation in Political Life: Norway and the
United States of America1, International Social Science Journal 12, 69-99.

Sainsbury, D. 1983. 'Functional Hypotheses of Party Decline: The Case of the Scandinavian
Social Democratic Parties', Scandinavian Political Studies 6, 241-260.

Statistisk tiårsoversigt, 1989. København: Danmarks Statistik.

Sundberg, J. 1987. 'Exploring the Basis of Declining Party Membership in Denmark: A
Scandinavian Comparison', Scandinavian Political Studies 10, 17-38.

Sundberg, J. 1989. 'Premisser för politisk massmedlemskab: partierna i Danmark i nordisk
jämförelse', Politico 21, 288-311.

Svensson, P. & Togeby, L. 1989. 'Græsrodsdeltagelse mellem politisk mobilisering og interessevaretagelse',
Elklit, J. & Tonsgaard, 0., eds., To Folketingsvalg, Århus: Politica.

Svensson, P. & Togeby, L. 1991. 'The Political Mobilization of the New Middle Class in
Denmark during the 1970's and 1980Y, West European Politics 14, 149-168.

Togeby, L. 1989. Ens og forskellig. Græsrodsdeltagelse i Norden. Århus: Politica.

Togeby, L. 1991. 'The Political Implications of the Increasing Number of Women in the
Labour Force'. Paper, Annual Scientific Meeting of The International Society of Political
Psychology, Helsinki.

Verba, S, & Nie, N. H. 1972. Participation in America. Political Democracy and Social
Equality. New York: Harper & Row.

Verba, S., Nie, N. H. & Kim, J. 1978. Participation and Political Equality. A Seven-Nation
Comparison. Cambridge: Cambridge University Press.