Voluntary associations and trust in Finland

The article is an empirical study of the hypothesis put forward in theories of social capital, according to which voluntary associations have a positive effect on trust in people. The central research question of the article is: are association memberships or active participation related to trust towards people in Finland? Looking at the number of association memberships and association registrations shows that association life has been active in Finland. The international comparison shows that particularly in the Nordic countries, people participate in associations on a vast scale. However, it is also important to note that while the old politicalideological associations have lost some of their importance in Finland, several new kinds of associations have emerged. Thus it is relevant to ask whether connections between trust and association activity vary between different types of associations. Empirical analysis shows that the members of only certain types of associations – charity associations, sports associations, and cultural associations – were found to be more trusting after various background variables had been elaborated. It was found that, in particular, the number of association memberships has a connection with a stronger level of trust. Even though memberships in associations have a statistically significant link to the trust that people experience, the percentages of variance explained are low.


Introduction
In Finland, the field of voluntary associations is still active.The Finnish association field is not in a crisis, as is shown by the numbers of association memberships and new associations founded.It is, however, known that the status of political associations has weakened.As early as in the 1970s, the field of class-based associations began to weaken.During the next decade, the influence of ideological divisions continued to dissolve, of which a good example, among other things, was the popularity of various hobby associations.This tendency was further strengthened in the 1990s.At the end of the decade, approximately three fifths of newly founded associations were various sports, culture, and leisure associations (Siisiäinen, 2002b).Recently, the association field has tended to specialize, which is evident from the sheer variety of associations -for example the multitude of new leisure associations.It has been said that these new associations are based more on consumerism and lifestyle, not so much on reaching wider social goals, which oftentimes require longstanding commitments (Alapuro & Siisiäinen, 2006).
Associations have been seen as mediating factors between individuals and the state.Associations draw volunteers together, therefore enabling them to further their interests, ful- The author works as a senior assistant at the Department of Social Sciences and Philosophy.Address: Department of Social Sciences and, Philosophy, P.O.Box 35 (MaB), FI-40014 University of Jyväskylä, Finland; E-mail: tomi.k.kankainen@jyu.fifil themselves and let their voices be heard.Associations also function as filters, gathering the various interests of citizens and fitting them to the level of society as a whole (Siisiäinen, 2000(Siisiäinen, , 1986)).Associations and organizations both channel and bring forward the ongoing interests and challenges of society.
Recently, the study of social capital has been active in Finland.Associations have been one of the interests of the study of social capital (Kankainen, 2007).Generally speaking, social capital refers to the communal characteristics of the social structure, such as social networks, norms and trust.These characteristics can promote interaction between the members of the network and the coordination of actions.Social capital (Gittel & Vidal, 1998;Putnam, 2000;Ruuskanen, 2003) has been analytically divided into three classes.Bonding social capital refers to the relationships between family members, friends and other such people, who are strongly committed to each other.Such networks, with strong bonds of trust and various norms, can be exceedingly solid.On the other hand, they can act exclusively towards other groups or individuals.Bridging social capital has to do with the weakest of social bonds -associations, for example, can act as sources of this type of social capital.In this case, the bonds between members of the network are not as binding or strong.With such weaker bonds, the creation of broader networks is easier (Granovetter, 1973), and tolerance towards different people can be stronger.This, in turn, can lead to cross-overs of different cultural and social boundaries.Linking social capital links together institutions and levels of functions, which differ from each other on a hierarchical level.

TOMI KANKAINEN
Studies suggest that the bonding social capital of Finns (the relationships between relatives, friends, and acquaintances), is weaker than in many other countries (see Lehtonen & Kääriäinen, 2005, 301).If this is indeed the case, the importance of Finnish associations becomes even stronger.According to studies, the action of associations is more important in Finland than in any other country: the number of associations and their members is high in Finland, and there are no well-functioning alternatives to them.This means that the changes in the action of associations and organizations are fundamental to Finnish society (Siisiäinen & Kankainen, 2009).
This article concentrates on studying the connection between associations and trust in Finland.Many international studies have presented the idea that associations are the central areas of society, through which trust is created (f.e.Putnam, 2000).It is important to ask which kinds of associations are prone to create trust in Finnish people.This will be one of the main issues in this paper.It is a vital question because the field of associations in Finland has experienced a fundamental change since the deep recession at the turn of the 1990s.

Changes in association activity in Finland and international comparison
Since Finnish independance, roughly 170 000 new associations have been founded in the country (Figure 1).Approximately 80 000 associations are currently active -it is impossible to come up with an exact figure.There is variation in the number of new registrations, but, generally speaking, the number has risen steadily since the 1920s.Putnam (2000) concern about the deterioration of the American association field cannot be applied with the situation in Finland.In the 1990s, more associations were formed than during any other decades.Figure 1 shows that during the time period examined, on average slightly below 2 000 new associations were founded each year.During the 1990s, the average was 2 600.In 1996-1998, the number of registrations was at an all time high (Siisiäinen, 2002b).Furthermore, the number of Finnish association memberships has definitely not decreased, rather, it has increased.The number of people not belonging to any associations has gradually decreased since the 1970s, from approximately 40 per cent of the population to less than 20 per cent (Kankainen, 2007, 73).
There has also been a clear internal change in the association field.Political and ideological associations have lost ground (Table 1) since the 1990s -a change which is also reflected among political parties.In the early 1960s, roughly one in five voters were members of a party, during the 1990s only a tenth of voters were party members (Sundberg, 1996).The turnout of voters has, some exceptions notwithstanding, decreased ever since the 1960s.The voter turnout of the parliamentary elections of 2007 was at its lowest since 1945.According to official member statistics, the parliamentary parties of Finland had approximately 360 000 members in 2004, which is almost half of what the number was during the golden era of memberships in the 1970s (Borg, 2006).Voters also are not as loyal to political parties as they used to be.Even though people usually vote the party for which they have previously voted, the voters' desire to highlight the significance of a candidate through their voting decision has increased.In the parliamentary election of 1991, over 40 percent of voters gave more importance to the candidate than the party in their own voting decision (Pesonen, 1991, 99-101;Pesonen, 1995, 116).In the parliamentary election of 2003, however, the number of people who considered the party more important than the candidate increased in comparison to the elections of the 1990s (Paloheimo, 2006).
While political and ideological associations have lost ground, exercise and sports clubs, cultural and various other leisure associations have increased their popularity (Table 1).In fact, 70 per cent of all associations registered during 2000-2002 are cultural, exercise or other hobby associations.This can be viewed as a growth in the percentage of lifestyle associations.The specialization of the associations establishment is evident in the large number of new associations for various sports and music associations, motor clubs and other hobby associations.It is not customary for such associations to try and reach wider social goals, which would also require a long-standing commitment.These associations usually offer ways of self-expression or self-development.
This view of the changes in the field of associations is further strengthened when the results of the Leisure studies of 1981, 1991and 2002by Statistics Finland are examined (Statistics Finland 2005).The participants of the studies were asked about their activities in associations during the last 12 months.In 2002, 52 per cent of Finns had participated in the activities of an association during the last 12 months.The percentage was exactly the same in 1991, and the result of the year 1981 does not differ significantly (57%).Men and women participated equally actively, while in 1981, men were more active than women.The most popular way of participating in associations was to take part in the activities of sports or exercise clubs.Such participation has also clearly become more popular.In 1981 17 per cent said they participated in the activities of sports organizations or exercise clubs.In 2002, 23 per cent participated.Men (25%) were slightly more eager to participate than women (20%).Sports organizations had gained popularity particularly among the young.During the same time period, the percentage of people participating in the activities of political parties had dropped from six per cent to two per cent.Participation in vocational associations had also sharply decreased: from 21 per cent to eight per cent (see Hanifi, 2007).
Behind the number of founded associations and association participation lie more common societal and cultural processes of change that can be seen in the changed methods of new associations: new associations form as highly specialised groups, among almost all kinds of associations.For example, the hobbies of small groups get their own associations, such as different forms and genres of music, various breeds of cats and dogs, sports and different forms of exercise, interests in various brands.Participation in associations of specialised interests in most cases means that a person  1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Table 1 Registrations of different types of associations in Finland, 1919-2002(Siisiäinen 2002a).
Type of association 1919-1969 1970-1979 1980-1989 1990-1994 1995-1999 2000- only invests a small part of one's personality and identity in association work.The other development is the increasing lightness and wildness of administrative structures.Thirdly, new associations are small, not only when considering the number of their members but also their budget and functional capacity.Studies have shown that small associations are more efficient in producing association activists; they are more capable than big associations in converting formal membership into functional activity.Nevertheless, the legitimisation of the political system and the functioning of the welfare regime require the existence and activity of volunteers and activists on a large scale (Siisiäinen & Kankainen, 2009).Socio-economical status has an effect on the participation in associations.In 2002, 65 per cent of upper managerial personnel had participated in the activities of at least one association.This was followed by 52 per cent of lower managerial personnel, about 49 per cent of production and service employees, 44 per cent of pensioners and 38 per cent of the unemployed had participated.65 per cent of agricultural entrepreneurs had participated in the activities of at least one association in 2002.While upper managerial personnel were more active association members than others, their percentage has decreased -in 1981, 76 per cent participated in activities of associations (cf.Hanifi, 2006).Education is also an important factor of social capital, according to studies.Education correlates with a higher level of trust towards people and various institutions.The highly educated also participate more actively in associations, they are more often in positions of trust.They vote more actively than others (Pääkkönen, 2006;Hanifi, 2006;Iisakka, 2006).The same applies internationally (Putnam, 2000).In this regard, the Finnish association field has become polarized and is in a crisis, as far as some groups are concerned; for example the majority of the unemployed do not participate in associations.Moreover, it appears that societal activity is "hereditary".Those Finns who feel that their parents have been active members of associations and interested in societal questions are members of more associations than others (Kankainen, 2007).
In international comparison, Finns are quite eager participants of associations, as far as association memberships are concerned (Table 2).The Scandinavian countries and the Netherlands resemble each other in citizen activity.In these countries, the association field is not showing any signs of deterioration and people are participating actively in associations (see Dekker & De Hart, 1999;Siisiäinen, 2000).In Denmark and Sweden, nine out of ten people are members of at least one association.Over 80 per cent of Norwegians and Dutch belonged to an association at the beginning of the millennium.Finland (76%), Austria, and Luxembourg closely follow.In Ireland, Belgium, Germany, and the United Kingdom two thirds of the population are members of at least one voluntary organization.In Eastern and Southern Europe such membership is significantly rarer (see Uslaner, 2003).It is also known that certain forms of social capital have accumulated geographically, per country.International comparisons (for example Kankainen, 2007) have shown that in countries where people have trust towards the functionality of democracy, as well as the schooling system and health care, the trust towards other people is also the strongest.Additionally, trust is strongest in countries where people are members of the highest number of different associations.In this sense, trust towards other people and institutions is connected with an active associations establishment.Even though Finns are, in the European context, members of several associations, Finland has also a peculiarity: when compared to the other Nordic countries, Finns are passive members of associations (op cit.).In comparison to the other Nordic peoples, Finns are less willing to participate actively in political groups.
As mentioned earlier, another Finnish peculiarity is that bonding social capital seems to be less common in Finland than in many other countries.In studies, bonding social capital has been measured by the number of relatives, friends, and acquaintances and by how often one meets with them.Such connections are particularly strong in Mediterranean countries.Finns have less bonding social capital than other Nordic peoples (see, for example, Lehtonen & Kääriäinen, 2005, 301).It would seem that the associations and the vol-

Voluntary associations and trust
In small communities, people know each other and are able to monitor each other.Among other things, the social order and stability of such communities is based on norms, the breaking of which leads to sanctions.Urbanization and individualization both raise questions of how to forecast and estimate the behaviour of unknown people.People's ability to make individual choices grew especially during the previous century.When globalization, changes in the movements of economic capital, and the uncertainty of employment are added to this, it becomes easy to understand that various themes of societal cohesion have become popular in both science and politics.Trust, networks and reciprocity have been put forward as central binding forces of human beings and societies.Trust has been thought of as the foundation of an effective economy and government (e.g.Putnam, 1993).
However, social capital and trust can have negative effects.Social capital can also be anti-democratic.People in certain groups and networks can concentrate on mutual interaction and maintain trusting and reciprocal relationships while simultaneously shutting everyone else out.In certain conditions, this may lead to accelerated polarization (e.g.Maloney, Graham, & Stoker, 2000).Somers (2005) remarked that if researchers are too attracted by the positive consequences of social capital and studying them, they may not notice processes, by which structures such as the welfare state are being stripped down; and how the equal rights of people are being dissolved via economic polarization.In Finland, it has been suggested that we should take more care of our fellow beings and participate in the activities of the third sector, so that people would be able to help each other through volunteer work.Some think that social capital is being applied as a conceptual apparatus, with which people are made aware of alternatives to the help and care services of the welfare state.
In international comparisons, the Finns are ranked as one of the most trusting peoples.This can be explained partly by how the educational system, together with other so called welfare services, works in a fairly just manner (Rothstein, 2002;Kumlin & Rohtstein, 2005).According to Putnam (1993;1995;2000), we know that associations are connected to social capital, which also acts as an externality on how trust becomes generalized.Putman's claims can be traced back to the classical ideas of sociological pluralism, which were presented by de Tocqueville in his work Democracy in America.Tocqueville (1951;2006) was concerned about individualism, which was a by-product of equality and changes in society.By individuality he referred to the tendency to spend one's life in a small circle of family and friends without any desire to influence and take responsibility of societal issues.This, in turn, leads to a concentration of power.Tocqueville famously found a solution to the problem of democracy and the fragmentation of power from local governments and associations.
Tocqueville emphasises the process of change which began in Europe during the early Middle Ages and which resulted in the replacement of the aristocratic system by democracy.The danger with democracy is, however, the political passivation and privatization of citizens.For Tocqueville, freedom meant that one trusts one's own thoughts yet is able to work in cooperation with other people in order to reach common goals.One does not learn such skills by interacting only with a closed circle of acquaintances; larger communities are necessary.It is therefore no surprise that Tocqueville saw associations as positive coalitions of interest, by which people could both strive towards their common goals and to learn skills that are necessary for freedom (Tocqueville, 1951;Saastamoinen, 1998;Siisiäinen, 1986).
Tocqueville sees associations as having numerous important functions in societies.Associations enable people to reach the goals they have set for themselves and to make their opinions heard (Tocqueville, 1951, 510).Additionally, associations make public the goals and targets people have.Therefore the freedom to form associations limits the number of underground societies.Thirdly, people meet and converse with each other in associations, which helps them to approve of and take into consideration differences in opinion (ibid.).Even though people may not exactly like each other, associations actually do further tolerance towards different opinions.Fourthly, associations can take care of matters which may be the government's responsibilities.For Tocqueville, this was crucial for societal peace.If people do not learn how to unite and resolve issues, the whole society would be under threat and in danger of succumbing to barbarity, according to Tocqueville (op cit., 503).And fifth, all activity in associations furthers the formation of political parties.The more people have various small common projects and coalitions, the better their ability to fulfil such projects (Tocqueville, 1951, 509;Ilmonen, 2006).
Tocqueville sowed the seeds of the central ideas of sociological pluralism and association theory.Simmel, among other classics of sociology, presented the idea that associations are the central areas of society, through which citizens' membership in a society is fulfilled.According to him, overlapping memberships also form solidarity between people from different social classes (Siisiäinen, 2006).At the end of the last millennium, Putnam (1993, 89-90;2000, 338), in particular, presented the topic of associations as types of schools of democracy.He makes a division between horizontal and vertical networks.Equality and a more or less equal division of power are typical to horizontal networks.Hierarchy and asymmetrical distribution of power is typical to vertical networks.Putnam's analytical division is viable, but in fact it seems to be that the types of networks are mixed.Nevertheless, Putnam favours horizontal networks over vertical ones.According to him, horizontal networks are the ones which particularly produce trust and interaction.He considers associations a prime example of such networks.When associations attract different people to work equally, the situation is beneficial for social capital.Thereby associations increase trust (Putnam, 1993, 173-4).For Putnam, associations as horizontal networks are central in the formation of social capital.However, this begs the question, to which extent associations actually are horizontal and open.At least some associations are quite hierarchical and exclusive about their membership.
Studies have shown that people with association memberships generally trust various institutions and people more than people who do not participate in associations (Putnam, 2000;Siisiäinen, 1999, 163-5;Ilmonen & Jokinen, 2002, 155-6).On the other hand, some studies have found that the connections between association activity and trust can be very weak (Dekker, 2004;Delhey & Newton, 2003).Because of this even the connections that have been discovered have been perceived as possibly stemming from selection (e.g.Stolle, 2001) -that people who are already trusting and fortunate tend to participate in associations.
In a study by Wollebaek and Selle (2002) it was not the association activity (time spent in associations) but the number of association memberships that furthered networking and trust in other people.Both passive and active people with several association memberships trusted other people equally (data: N=1695; 16-85-year-old Norwegians).This suggests that the view Putnam has on the significance of dense interaction may be an exaggeration.Mario Dian's studies ( 2004) suggest the same.He studied the members of associations in Lombardy (N=1091).People who have invested less time in an association, according the study, trust political officials somewhat more than others.Stolle (2003) aimed to study the influences of association memberships by trying to eliminate the effect of selection.His aim was to examine whether the length of time a person has been a member affected trust and citizen activity.The data consists of a thousand members of various associations from Sweden, Germany, and the United States.The study could not confirm the claim that associations do generalize trust outside the associations.We should also remember, as Kwon (2004) in his insightful article emphasizes that social capital was used as the motor of fascism just as effectively as it has functioned as the cornerstone of democracy.Hence, associations can cause discord (Green, Preston, & Germen, 2006, 177) as much they can connect different people.

Research questions, data and measures
Activity in associations has been thought to create a sort of "school of democracy", where citizens can voice their opinions, make their goals public, meet with each other and discuss, thereby learning to accept and to take different opinions into consideration (see Tocqueville, 1951).Even though people might not necessarily like each other, associations help raise tolerance towards different things.Tocqueville considers this necessary to a peaceful society.The more people join together and create small co-operative projects, the better they can carry them out.These processes join citizens of different social classes and layers, help to strengthen solidarity and increase trust between various layers.
If associations affect trust on an individual level, the trust of those belonging to associations needs to be stronger than that of other groups (see appendix 1).If active participation in associations is relevant to trust, the active members of associations should have more trust than others.Also, as previously noted, a presumption exists that several intertwining memberships were particularly effective in generating trust.Therefore, the more association memberships an interviewee has, the more he or she should have trust (see Table 4).It is also important to note that while the old political-ideological associations have lost some of their importance in Finland, several new kinds of associations have emerged.Thus it is important to ask how the members of these new types of associations trust familiar and strange people in comparison to the members of old ideological and political associations.In conclusion, it can be said that the central research problems of the article are: 1. Are association membership or active participation related to trust towards people?

Do connections between trust and association activity vary between different types of associations?
The quantitative data on which the analysis is based was collected during March and April in 2005.1000 Finns answered the questionnaire.The data was gathered by computerassisted telephone interviews using both wired and mobile phones.The sample was random.If an interviewee refused to be interviewed, a random backup person was selected.The data is representative of the population in terms of gender and age.The data is somewhat selective in terms of professional education.The interviewees are slightly better educated than an average Finn.Secondly, the unemployed are under-represented.Over 7 per cent of Finns were unemployed in 2005, whereas 4 per cent of the data consists of unemployed Finns (see Kankainen, 2007).
The results of this study are confirmed by the fact that the analysis uses and compares data from numerous sources.In addition to the actual main data, I have also used the Leisure survey conducted by Statistics Finland in 2002-2003 for comparison (see f.e.Iisakka, 2006).3355 Finns took part in this survey.The response percentage was 73.The data was gathered by visiting the respondents and with the additional assistance of a computer interviewing them.The third data used is the data from the "Economic depression and trust" research project, which interviewed a total of 1717 Finns after one repeat questionnaire (Ilmonen & Jokinen, 2002).Its response percentage was 49.The fourth data is the International Social Survey Programme data from the year 2004 (N=1247).Only the answers of Finns have been used in this data.This article uses analysis of variance and regression analysis from multiple variable methods.Their statistical presumptions (such as normal distribution, homoschedasticity and non-multicollinearity) are valid.
Dependent variables are various variables which measure trust.It has been usual for studies on trust to ask only about generalized trust without a specific target and context.This has been a flaw in many previous empirical trust studies.Now it is possible to examine the trust people have towards each other better than before, using data gathered in 2005.First of all, trust was measured using the claim "Generally speaking, one can trust other people" on a scale of 1-5 (complete disagreement -complete agreement).Secondly, people were asked how much they trusted their neighbours, unknown Finns, and foreign persons by asking them to "Tell how much you trust the people/groups listed below" on a scale of 1-5 (not at all -very much).Thirdly, the following situation was presented: assume that you are sitting in a café and you have to leave your luggage in someone else's care for a while.In this situation, how much would you trust a) your neighbour, b) unknown Finns, and c) unknown foreigners on a scale of 1-5 (not at all -very much).In some cases sum variables were used as indicators of trust in neighbours and unknown Finns.Validity has been of most importance in the creation of the sum variables, but their statistical viability is measured by the Cronbach´s alpha, which measures the inner consistency of the sum variable.The Neighbours sum variable (Cronbach's alpha=0.71) was created by calculating the mean of the answers to the questions about trust in neigh-bours generally and in the café context.The Unknown Finn sum variable (Cronbach's alpha=0.74) was created by calculating the mean of the answers to the questions about trust in unknown Finns generally and in the café context.The Unknown foreigner sum variable (Cronbach's alpha=0.76) was created, in turn, by calculating the mean of the answers to the questions about trust in unknown foreigners generally and in the café context.
Besides association memberships, the independent variables are education (no professional education, or vocational school; folk high school level, tertiary education), income (low income below €2000; middle income €2000-2999; high income over €3000), sex, age, optimism and the population of home municipality.Because education and income were clearly measured by ordinal scales, they were used as dummy variables (those with low income and the least education were made the reference groups).Studies suggest that those of high societal status tend to be more active in associations (e.g.Putnam, 2000, 62-3).It is known that education and income correlate with a higher sense of trust.Hence it is necessary to elaborate the connections between association activity and trust in other people with different background variables.As mentioned, such standardizing independent variables in this article are education, income, age, and gender.Additionally, optimism may show how well people are doing.In this study, optimism has been measured with the claim "I believe that the future holds many good things," to which the interviewee has been asked to answer on a scale of 1-5 (from "I disagree completely" to "I agree completely").Information about association memberships was gathered by presenting different types of associations and asking the interviewees to tell if they were members of the associations.The variable of the number of association members has been created by counting the number of associations the interviewees belonged to.The activity in associations was measured by asking how often the interviewees had taken part in the activities of each association during last 12 months.

Empirical analyses
The connections are examined (Appendix 1) by comparing mean values, which, in addition to the differences, also shows the levels of trust in different groups (Anova, Tukey post-hoc test).Those who have participated in association activities once (or less) during the previous year have been categorized as passive members.The rest are active members. 1 The connections between trust and association membership have been elaborated on the basis of income and education.This applies to all membership types.
The first observation is that belonging to associations is not necessary equal to higher trust.The trust in unknown Finns and foreigners is especially strong among those who belong to a charity association (Appendix 1).It is interesting that the passive members of charity associations in particular are more trusting.Passive members of charity associations have more trust in unknown Finns in the café context (p<0.05)than people who are not members of charity associations.Secondly, they have more trust in unknown foreigners in general (p<0.01) and in the café context (p<0.001)than those who are members of such associations.
Previous studies have observed that sport and exercise clubs, cultural associations and various others associations for leisure hobbies have become more popular in Finland (Siisiäinen, 2002b).Members of these associations would also seem to be more trusting towards people.In this sense social capital is not threatened in Finland.Members of sports clubs have more trust towards Finns in general (p<0.05) and in the café context (p<0.05)than non-members.Secondly, they have more trust in unknown foreigners in cafés (p<0.05)than others have.The mutual comparison of these groups shows that active members of sports clubs in particular are more trusting towards unknown people.Seippel (2006) has previously noticed in his study of Norwegian members of sports clubs that the sports club members trust people generally somewhat more than the population on average.
Belonging to associations for hobbies and cultural associations also correlates with a stronger sense of trust in unknown other people.People in such associations have more trust in Finns in general (p<0.05) and in the café context (p<0.01)than non-members.A closer examination shows that especially active membership in such associations is linked with trust (Appendix 1).Passive membership in retirement associations, on the other hand, correlates with a lower level of trust.
The trust that active members of international friendship associations show towards unknown foreigners is strong in comparison to others.The examination of mean values also shows that members of other economic or professional associations have more trust in unknown foreigners in general and towards both Finns and foreigners in the café context when compared to non-members.Standardizing education and income, however, diminishes these correlations.Active members of youth associations also show strong trust in unknown people, but because there are so few of them in the data, the differences are statistically significant only in the trust they show towards unknown foreigners in cafés (p<0.05).
Previous studies have shown that Finns trust their family and friends almost without exception.Many Finns also trust their neighbours, co-workers, classmates and superiors.(Kankainen, 2007;Ilmonen & Jokinen, 2002).Therefore, it is not surprising that membership in associations and active participation in them does not correlate with trust in neighbours; active members of cultural associations or associations for hobbies and other economic or professional associations form an exception.
Passive membership in retirement associations, however, correlates with a lower level of trust.The "Economic de-pression and trust" data 2 and the ISSP data 3 support the aforementioned results.Trust and membership in associations can also be examined in the Leisure survey of Statistics Finland (Appendix 2).Again, the results support the aforementioned findings.In this survey, trust has been examined by using the claim "Generally speaking, people can be trusted".According to the study, 84 per cent of Finns trust people in general.First of all, it can be seen that members of sports clubs have more trust towards other people on average than non-members (p<0,001).Secondly, membership in professional associations would also seem to correlate with a stronger level of trust (p<0.01).Thirdly, members of friendship associations or other international associations trust people more than others do (p<0.05).Fourthly, the trust among cultural and art associations is also strong -even though the difference is not statistically significant.Again, the members of retirement associations show a lower level of trust than others (p<0.05).
In Finland, it would seem that only certain types of associations are correlated with trust.Sports clubs and cultural associations, which have increased numerically, are also strong in terms of trust.People participating actively in these associations trust other people especially.Emphasizing the role of association memberships in the Putnamian fashion might, however, be an exaggeration.For example, it is specifically the passive members of charity organizations who show most trust in other people.Additionally, the data of Statistics Finland offers preliminary support to the claim presented in theories of social capital, according to which active participation in associations and the bonds created therein can strengthen the boundary between members of the group and people outside of the group.Activity in associations can strengthen the bonding between people who know each other, but at the same time weaken the bridging of social capital -for example trust in unknown people.In this study, people participating in religious associations, circles and congregational clubs at least twice a week (N=40) had a generalized trust which was lower than the average (Anova, p<0.01).This empirically suggests that associations do not necessary generalize trust among citizens.
Let us try to clarify the numbers by taking a look at memberships in different types of associations.If we assume that associations produce trust and heighten tolerance towards unknown other people, this might stem from overlapping memberships in several associations.Such a person would be more likely to meet different people through the associations.For example, in the study conducted by Wollebaek and Selle (2002) the trust towards other people and networking were enhanced by the number of association memberships instead of activity in associations.The more memberships in different types of associations Finns have, the more trust they also show towards their neighbours and towards both unknown Finns and unknown foreigners (Table 3).This correlation can also be thought of as showing that the more people trust other people, the more likely they are to be members of associations.It is very difficult to show the causal direction of the correlation.
Social capital -whether it is measured as trust or as mem-berships in associations -has a tendency to accumulate for a certain type of people.According to earlier studies, education is very central in explaining social capital in Finland (Iisakka, 2006, 106).This article also supports the hypotheses that trust, human/cultural capital and economical capital are connected to each other (Table 4).Those of a lesser education and low income have less trust towards people than others.However, the correlations between the number of memberships and the amount of trust remain when the effect of different background factors is standardized in the regression analysis (Table 4).The general trust alone (generally speaking, people cannot be trusted) does not correlate with the number of association memberships in Finland.The coefficients of determination are, however, very small.My data supports the idea that membership in associations in Finland does have a connection with trust, but that its influence should not be overestimated.

Conclusion and discussion
The Finnish association field does not seem to be in any sort of crisis.It has, however, gone through an internal change.The flow of active association members has moved away from political and ideological associations, towards sports, culture and leisure associations.Even though a general crisis is not to be seen, and challenges and problems abound.In addition to a some type of crisis on the field of political activity, the unemployed and the marginalized are not only members of fewer associations, they are also more passive members than others.They also participate more seldom in volunteer work than others.Activity in associations and volunteering is connected to the socio-economical status and the education of a person.The different capitals (economical, cultural, and social) seem to accumulate to certain groups of people (Siisiäinen, 2000;Sanaksenaho, 2006).
It has been brought up in public discussion that people should take more care of their fellow beings and to participate in the activities of the third sector.The communitaristic 2 According to the Economic depression and trust data (2000-2001, N=1717), people who belong to other economic or professional associations (Anova, p<0.05) or an association that is related to studying (Anova, p<0.01) trust unknown Finns more than others do.Members of sports clubs (p<0.01),cultural associations (Anova, p<0.05), charity associations (p<0.01),other economic or professional associations (Anova, p<0.01), and associations related to studying (Anova, p<0.01) had more trust in unknown foreigners than others had.Passive membership in retirement associations has a connection with a lower level of trust (Anova, p<0.05) in this study, too.
3 The ISSP data (2004; N=1247) enables the study of the connections between generalized trust and membership to political parties, professional or economic associations, sports clubs, cultural or other leisure associations, and churches or other religious associations.Especially those who belonged to sports clubs, cultural associations or some other leisure association (Anova, p<0.001), and also those belonging to economic or professional associations (Anova, p<0,05) were more than the average of the opinion that generally people can be trusted.The ISSP study asked about memberships of only the types of associations I have mentioned.

Table 3
The correlation of trust in neighbors, unknown Finns and foreigners and generalized trust between the amount of association memberships (%).idea of a sense of community and the help of near relations is, understandably, a charming and romantic one, but it is also problematic.The situation is paradoxical in the light studies -those who most need the support of various networks are those who have the least social capital.It is therefore unlikely that the support of near relations and the third sector would be enough to mend the polarization caused by the weakening of the welfare state institutions (c.f Somers, 2005, 5-7).This study has examined the hypothesis put forth in theories of social capital, according to which associations have a positive effect on trust in people.Special interest was shown towards the trust people felt towards unknown other people.Through the use of several data it was possible to point out that the members of only certain types of associations -charity associations, sports associations and cultural associations -were found to be more trusting after various background variables had been elaborated.In particular, the number of association memberships has a connection to a stronger sense of trust.Those who were members of several different associations trusted unknown Finns and foreigners more than others did.However, my analysis systematically shows that the connections between associations and trust are weak.Undeniably, associations have had many crucial roles in the Finnish society, for example in creating welfare state, but there are no reasons to overestimate the role of associations in enhancing trust (see Dekker, 2004;Delhey & New-ton, 2003;Stolle, 2001).
Even though the effects seem minor on an individual level, associations have probably had a central affect on the strong Finnish culture of trust.This is because associations have participated in creating a welfare society with its various institutions.Studies show that fairly functioning welfare services -such as the Finnish educational system -promote trust between people (Kankainen, 2007;Rothstein, 2002).Also, as previously mentioned, countries with a high number of association memberships also have a somewhat higher level of trust towards other people, the parliament and the justice system, and people are satisfied with the democratic process, the educational system and healthcare.It is possible that an active association field and different institutions work (when considering trust) to reinforce each other.When institutions function in a just and professional manner, they create the kind of culture of trust that can help and activate people to participate in societal activity.However, associations can simultaneously have a central position, as they can present development suggestions and critical points of view via their large numbers.
Passive members of associations are not, however, categorically more distrusting than others.Therefore, as far as trust is concerned, passivity cannot be seen as a negative phenomenon in the way Putnam has claimed.This does, however, beg the question: how should the aforementioned be viewed?One possibility is that trust may make people passive.By this I mean that these passive members of several associations trust unknown other people, but they also trust that the people who are active in the associations do their job well (Simmel, 1908;Richter, 1985).By being members of the associations these passive people show their support to certain things and also join a certain kind of community.For example, the people in charity associations are, by and large, passive, but also very trusting.Perhaps, generally speaking, these people tend to show a kind of humanistic trust, which then manifests a strong trust in unknown other people and institutions.
However, the less we meet different people and listen to their thoughts the more likely we start to draw lines within society and more likely society will polarize.Associations are central to society where the citizens' membership in a total society is fulfilled, according to Georg Simmel.Various overlapping memberships may also create solidarity and trust between people of different social classes.The traditional Finnish school lunch situation can be considered as creating these effects as well as any; during lunch young people with different backgrounds are having their meals together.As long as Finns wish to preserve their universalistic politics, it is easier for people from more modest economic backgrounds to study and be active in many ways, which does not, at least negatively, affect the solidarity and trust between citizens.

Table 2
Memberships of associations and voluntary organizations per country (%).
* European Social Survey data 2002 (weighted); association data does not exist for some countries * the countries are listed in the order of trust experienced towards people, most to least unteer work done within associations are especially important in Finland, because there are fewer other channels of social capital.It has also been thought that participation in associations does not only create trust between association members, but that it also strengthens trust towards other, unknown people.

Table 4
Regression analyses (OLS) of trust towards other people in Finland (n=950).
APPENDIX 1 Comparing trust in different types of associations by membership and activity (mean values, data 2005) Leisure survey conducted by Statistics Finland in 2002-2003 Statistically significant differences have been underlined (Anova, tukey post-hoc-test)