Generations and Socialization into Electoral Participation in Finland

ISSN 0080 – 6757 © 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

Generations and Socialization into Electoral Participation in Finland 1ORIGINAL 30 © 2007 The Author(s) ARTICLE Journal © 2007 Nordic Political Science Association Blackwell Oxford, Scandinavian SCPS 0080-6757 Nordic UK Publishing Political Political Ltd Studies Sciencecompilation Association

Hanna Wass*

This article examines the extent to which political socialization accounts for generational differences in electoral participation found in recent studies. Political socialization is defined as the learning process in which an individual adopts various political attitudes, values and patterns of actions from his or her environment. The analysis is based on the Finnish National Elections Study 2003. The results show that even though politics has had the smallest role during the formative years of the youngest generation and they most often do not know their parents’ partisanship, this generation has received the most encouragement for voting and the attitudinal change towards voting within an individual’s life span has been the most positive. Consequently, the study shows that if there were no differences in the socialization between the youngest and the older generations, the difference in turnout would be larger if only sex and socioeconomic factors were taken into account. Based on these results, the author draws the conclusion that, rather than political socialization, the factors behind the low turnout among the young generation have to be searched for elsewhere.

Introduction Students of political science have long been aware that electoral participation is strongly connected to age as well as to socioeconomic and attitudinal factors. Electoral turnout has traditionally resembled the curve of a semi-circle: it is lowest at the beginning of one’s adult life, rises with age and stays at the same level during middle age, and then gradually declines when a person grows old and faces physical infirmity (see, e.g. Tingsten 1937; for an overview, see Milbrath 1965, 134 –5). According to the explanation based on an individual’s life cycle (see, e.g. Campbell et al. 1960), low turnout among young age cohorts is a transitory phenomenon as it usually rises when a person grows older and adopts several ‘adult roles’ (for a general overview, see Highton & Wolfinger 2001, 202–3). Some recent studies, however, have shown that there is a clear generational effect along with life cycle and period effects (Perea 1999; Lyons & * Hanna Wass, Doctoral Candidate, Department of Political Science, University of Helsinki, PO Box 54, F-00014 Helsinki, Finland. E-mail: [email protected] Scandinavian Political Studies, Vol. 30 – No. 1, 2007

1

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

Alexander 2000; Blais et al. 2004; Franklin 2004; Franklin et al. 2004; for Finland, see Martikainen & Wass 2002; Wass 2006).1 Unlike the life cycle approach, the generational explanation claims that abstention has, to a certain extent, become a permanent feature in the political orientation of young age cohorts. This explanation is compatible with the paradox or ‘puzzle’ of declining turnout already noted over twenty years ago (see Brody 1978; Abramson & Aldrich 1982). Although the young age cohorts are better educated than their predecessors, overall turnout is declining. Moreover, it has been shown that even while holding socioeconomic factors such as education, class, income and housing tenure constant, the lower turnout of young voters remains unexplained (Martikainen et al. 2005). As an individual’s socioeconomic status usually rises with age, and those with a higher status are more likely to vote than those with a lower one, the independent effect of age to turnout suggests that the life cycle explanation is not sufficient on its own. Which factors account for this generational effect in turnout? Perea maintains that the tendency of younger age cohorts to vote less than their predecessors does not stem from, for example, their lack of political experience, but ‘rather by the fact that they belong to a generation that does not attach enough importance to the electoral process, or feels excluded or alienated from politics, in part due to a particularly demobilizatory socializing process common to the whole cohort’ (Perea 1999, 28; emphasis added). Consequently, as there certainly are several other possible explanations for generational differences in turnout such as the effect of younger enfranchisement (Franklin 2004) and sociodemographic, attitudinal and contextual factors and factors related to political interest and information (Rubenson et al. 2004), it seems reasonable to assume that it could be explained partly by the generational differences in political socialization. By ‘political socialization’, I mean the learning process in which an individual adopts various political attitudes, values and patterns of actions from his or her environment. Moreover, studying the role of political socialization is particularly important as studies in the field have been relatively rare lately. In the Mannheimian theoretical tradition, it is thought that a generation is formed by certain ‘key experiences’ that occur during its members’ formative years (see, e.g. Delli Carpini 1986, 8 – 9).2 Therefore, it is important to detect which factors in the socialization process of the young age cohorts have caused them to find voting a not very appealing form of political participation. In this article, the question will be approached by examining the differences between four generations in their socialization into electoral participation. It should be pointed out that, on the one hand, political socialization functions as a framework for key experiences as it has an effect on the impact of those experiences. On the other hand, political socialization may form a key experience by definition. These are quite complicated issues, but the main standpoint of this article is the idea that during a particular period 2

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

of time, a particular socialization atmosphere has prevailed and it has had a significant impact on various generations’ electoral participation (see Grönlund et al. 2005a). After discussing the previous research in the field of political socialization, I will first look at early socialization taking place in adolescence. I will analyse the kind of role that politics has had in the childhood of different generations, the amount of encouragement for voting generations have received from their parents and whether they know what party their parents voted for during their adolescence. Second, attitudinal changes towards electoral participation during an individual’s life cycle will be studied. Finally, I will examine how much the differences in turnout between the four generations examined can be accounted for by the aforementioned socialization factors. The analyses are based on the Finnish National Election Study 2003 (N = 1,270).

The Concept of Political Socialization The term ‘political socialization’ first appeared in political science in 1959 when Political Socialization by Herbert Hyman was published. Hyman defined the concept as the learning of social patterns mediated by various socialization agencies to whose social positions this learning process corresponds (Hyman 1959, 25; see also Dudlay & Gitelson 2002, 175). In the study of political socialization, family, school, friends and the media have usually been considered the most salient agencies of socialization (Ljungberg 2003). In his study of research tradition and future possibilities of research into political socialization, Renshon (1992, 443) suggests that the attraction to the field of political socialization in the 1960s is due largely to its compatibility with the behaviouralist approach that dominated political science at that time. According to Dahl (1961), behaviouralism, emulating natural science, emphasized four major points in studying the political process – all of which could be met by examining political socialization. First, the importance of an individual for the functioning of political institutions and processes became acknowledged. Second, the behaviouralist approach emphasized the importance of developing an interdisciplinary political theory. Third, systematic measurement strategies were accentuated, and finally, the development of generalizable theories regarding political behaviour and its causes was adopted as a goal (Renshon 1992, 443). Sapiro (2004, 2) states that the most significant theoretical justification for the study of political socialization is found most explicitly in Almond and Verba’s Civic Culture (1963) and in Easton’s system theory (1965, 1967). At the macro level, political socialization was seen as a crucial mechanism for the development of political culture (Almond & Verba) that enables democratic institutions and practices to function, or for the development of a diffuse Scandinavian Political Studies, Vol. 30 – No. 1, 2007

3

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

support (Easton) of the political system. At the micro level, research into political socialization concentrated on patterns and processes with which individuals engage during their development and learning processes, and that negotiate their relationship to the political context around them. While these two approaches to political socialization are complementary, they have directed the research questions and methods, to a certain extent, into different directions as those orientated towards the micro level have concentrated more on developmental and cognitive processes (Sapiro 2004, 2–3). However, the concept of diffuse support has been, at least to some degree, able to bridge the gap between macro- and micro-level approaches (Dudlay & Gitelson 2002, 176). In contrast to specific support, diffuse support is not based on the output of the political system, but on the political attitudes and values citizens hold towards the prevailing system and is thus essential for its stability. Childhood was viewed as a basis for diffuse support and, consequently, much of the early work concentrated on children (Dudlay & Gitelson 2002, 176). The study of political socialization flourished during the entire decade of the 1960s. In effect, it was diversified to such a degree that in 1970 Greenstein declared that: ‘Political socialization is a growth stock.’ According to Greenstein (1970, 970–2), one could indicate at least four prevailing definitions of the concept of political socialization: the study of children’s political orientations; the study of acquisition of prevailing norms; the study of any political learning, whether of conformity or deviance, at any stage of an individual’s life cycle; and actual observations of socialization processes, in any aforementioned meaning, taking into account both the socialized and the socialization agents. Hence, it became obvious that the field needed some clarification. In the late 1970s, the study of political socialization began to move into a new direction characterized by three interrelated foci (Niemi & Sobieszek 1977, 210–11). First of all, scholars in the field started to pay more attention to the adult lifespan, recognizing that adult attitudes may change more substantially than was earlier acknowledged. The second focus concerned the importance of the life cycle, period effects and generational effects in explaining that change. The third focus was related to the first two as it emphasized the role of events and environment in general as socialization agents. However, from the end of the 1970s, research into political socialization declined markedly (see, e.g. Dudlay & Gitelson 2002, 176) and the topic is nowadays much more popular among educationalists than scholars in political science (Borg 2004, 76 –7). Post-modernization theories aimed at describing the changes Western societies were facing gained ground. These theories emphasized, among other things, the value shift from materialist to post-materialist values (Inglehart 1977), the loosening of traditional collective ties and thus general individualization. The role of an individual was seen as 4

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

central in defining his or her identity, and it was recognized that this identity is not necessarily stable, but can be negotiable depending on the context. This discussion was to an extent contradicting the premises of the study of political socialization, according to which political behaviour is considered to be the result of a learning process in which the individual adopts different influences from the outside.3

Debates on Political Socialization The study of political socialization has also faced some critique (see Greenstein 1970, 273–6). According to Greenstein, it has been claimed that political socialization studies have a status quo bias. They are capable of explaining the processes of pattern-maintenance, but not those of change. This criticism is directed towards the branch that concentrates on studying how people absorb prevailing norms. However, Greenstein suggests that, in addition to the fact that prevailing norms may be absorbed, it is also possible to learn to reject them. Therefore, a person’s behaviour can be similar or dissimilar to his or her own generation or to the preceding generations. Moreover, at the level of political processes and institutions, the effect of an individual’s behaviour, whether showing continuity or not, may lead either to political continuity or change. In either case, behaviour is in part explained by situational stimuli and in part by a person’s attitudes and values, the latter of which is undeniably learned or ‘socialized’. In effect, some changes in political involvement and participation may originate from generational replacement due to the fact that the young cohorts have adopted essentially different values and patterns of behaviour to the previous cohorts and keep them throughout their life cycle (see Niemi & Sobieszek 1977, 226). As a matter of fact, this divergence actually makes them a generation. Although divergent development of the younger generation does not necessarily (or not even generally) lead to generational cleavages, some potential for this still exists (Niemi & Sobieszek 1977, 224). It is essential to recognize this potential for a ‘clash’ between generations, given that the Mannheimian model of generations can be considered foremost conflict theory as generations clearly differ from each other due to different key experiences and their concomitant impact on the attitudes and values of that generation. However, due to the conservative connotations of terms such as ‘education’, ‘socializing’ and ‘socialization’, it has become more popular to talk about, for example, ‘empowerment’, although the phenomenon under examination has been the same (Borg 2004). In addition to having conservative bias, the study of political socialization can be criticized for being normative. If we think, for example, about socialization into electoral participation, we may ask what the ‘goal’ for a successful Scandinavian Political Studies, Vol. 30 – No. 1, 2007

5

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

socialization process is and whose values this process ought to project. Various theories of democracy present very differently the ideal amount of participation among citizens, on the one hand, and the ideal scope of participation, on the other (see, e.g. Dalton 1988; Setälä 2003). Consequently, political socialization may very well be used as an explanatory variable while examining the causes of differences in electoral participation between generations, but we should be careful not to load any prescriptive value onto it. To a certain extent, political socialization is always the product of a certain period: while political participation was more or less considered to be voting during the 1950s and 1960s, the action repertoire started to expand heavily during the 1970s and has widened ever since (see, e.g. Van Deth 2001). It might be the case that because of this widening, the importance of voting as a form of political participation has been emphasized less during the formative years of the youngest age cohorts than before.

The Classification of Generations The categorization of generations in this article is originally from J. P. Roos (1987), based on data of Finnish autobiographies. While this classification may certainly be disputable, it is unquestionably the most widely known description of Finnish society. Apart from that, there are two other reasons why this categorization is chosen instead of, for instance, 10-year cohorts. First, it is much easier to interpret the relationship between socializationrelated factors and generation as we already have some idea of the identities and key experiences of generations under investigation. Second, the generations based on this categorization are large enough to gain statistically significant results given that the study is based on survey data with a limited number of observations (N = 735 for the most of the variables, see next section). By applying and supplementing this categorization, it is possible to identify four current Finnish generations: those born in 1900–1939, the generation of war and reconstruction; those born in 1940–1950, the generation of transformation; those born in 1960 –1969, the generation of the suburban and those born from 1970 onwards, the generation of individual choice (for the term, see Purhonen 2002). The key experiences for the first generation have been poverty, insecurity and illness, disintegration of families due to death of parents, short education and early entering into working life. However, it has undergone some prosperity after the reconstruction. During the parliamentary elections of 2003, members of this generation were 64 years old or older. Compared to the oldest generations, the life of the generation of transformation (44–63-year-olds in 2003) has been essentially different. Most of them were born in the countryside and moved to cities as adults, were educated longer than their ancestors and became wealthier. The generation of 6

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

the suburban (34 – 44-year-olds in 2003) also has its distinctive features. As a matter a fact, their life has been carefree to the extent that it could be stated that their key experiences have been certain ‘non-experiences’: their lives have been relatively steady and problems have been small.4 However, this generation has faced the worst economic recession of the postwar period while being young adults in the beginning of the 1990s. The youngest generation – those born in the 1970s and the beginning of the 1980s (18–33-year-olds during the parliamentary elections of 2003) – is more difficult to identify. According to Purhonen (2002, 15), they could be called the generation of individual choice. Without delving deeper into this definition, we can list some of the probable key experiences for this generation such as the economic recession in the beginning of 1990s and Finland’s accession to the European Union and globalization in general. In the youth barometer of 2004, a partly similar age group (15–29) most often called themselves the ‘IT-generation’ and the ‘mobile-generation’ referring to the expansion of new technology (Wilska 2004, 102–3). On the one hand, the unifying feature of this generation may be the lack of common experiences due to general individualization (see Purhonen 2002, 15). On the other hand, the inability to name key experiences can be caused by temporal closeness as this generation has existed for a considerably shorter period than the previous ones. A summary of the key experiences of each generation is presented in Table 1.

Table 1. The Key Experiences of Finnish Generations Year of birth

Age in 2003

The generation of war and reconstruction

1900–1939

64+

The generation of transformation

1940–1959

44–63

The generation of the suburban

1960–1969

34–43

The generation of individual choice

1970–

18–33

Generation

Key experiences War, want, the lack of or interrupted education, early entering into working life, scarcity, rationing, gradual increment of prosperity due to reconstruction Industrialization, urbanization, broadening of education, rapid increases in standard of living ‘Non-experiences’, the lack of experience of illness, disasters and want, smoothness of life yet recession during early adulthood General individualization, globalization, the acceleration of technological development, competition, consumption, emphasized role of the media

Sources: Roos 1987; Purhonen 2002 (quoting Hochschild, the term ‘the generation of individual choice’). Scandinavian Political Studies, Vol. 30 – No. 1, 2007

7

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

Data and Statistical Analysis The analysis is based on the Finnish National Election Study 2003. The data were collected via face-to-face interviews (N = 1,270). The respondents were also asked to fill in a supplementary, self-administered questionnaire (N = 735). The analysis presented here is mostly based on the latter part of the data except for the variable concerning the knowledge of respondent’s parents’ choice of party, which is from the first part of the data. The first part of the analysis is descriptive based on cross-tabulations. The role of politics in the respondent’s childhood and adolescence; the amount the respondent received advice about the importance of voting from his or her parents; the respondent’s knowledge of his or her mother’s and father’s choice of a party; and the respondent’s attitudinal change towards voting during their life course were all cross-tabulated according to generation. The purpose of the cross-tabulations is to examine the differences between four different generations in their socialization processes. The second part of the analysis is explanatory. I will investigate the impact of these four socializationrelated variables on the relationship between generation and turnout (i.e. will the differences in turnout between generations remain after taking into account the differences in their socialization processes). This was done by using logistic regression with the first category of each variable used as the reference category. The dependent variable – turnout – was weighted separately for men and women in order to correspond to the official turnout from electoral wards for the parliamentary elections of 2003. This was necessary as people have a tendency to over-report their electoral participation in surveys (see, e.g. Martikainen & Yrjönen 1991, 82; Pesonen et al. 1993, 531, for Finland). Results are presented in terms of odds ratios (ORs). ‘Odds’ means the probability to vote divided by the probability of not voting. Correspondingly, ‘the odds ratio’ is defined as the ratio of odds in a certain category of an explanatory variable to the odds in the reference category: ORi = (pi/(1 − pi))/(p0/(1 − p0)), where pi is the probability of voting in the ith category of an explanatory variable and p0 is the probability of voting in the reference category.

Empirical Analysis Socialization during Childhood and Adolescence Traditionally, the childhood family has been seen as the primary source of values and attitudes. However, this kind of ‘inject’ model has been re-examined 8

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

and it has been shown that different generations have developed their opinions horizontally rather than vertically from a parent to child due to similar experiences (Connell 1972). Also, others (e.g. Friedman et al. 1972) have noticed that the correlation in social and political values between parents and children is low. When it comes to political engagement, the socialization background, as measured by parents’ education and occupation, does not seem to have a large impact (Jääsaari & Martikainen 1991). On the other hand, the survey study by Jääsaari and Martikainen (1991, 42) shows that parents’ electoral participation has a clear independent effect on youth’s political interest. Consequently, it seems quite obvious that the environment where an individual grows up, consisting of family, school and various other social networks, must have some role in shaping the individual’s image of society and his or her place in it. The respondents’ views of the role of politics during childhood and adolescence are presented in Table 2. Most of the youth of today, the generation of individual choice, are the children of the generation of transformation. As the latter grew up in an atmosphere of active youth politics, it can be assumed that this activism has been visible in the childhood of their children. Table 2 shows, however, that politics has clearly had the smallest role in the youngest generation’s childhood and adolescence. Instead, politics has played a substantially larger role in the childhood and the adolescence of the two middle generations. In this respect, it seems very interesting that only a quarter of the generation of war and reconstruction consider the role of politics to be large or relatively large. Apparently, experiences of war have not been considered political, but rather perceived mostly as part of day-today survival. Moreover, a certain political prudence was distinctive to the postwar atmosphere, which, at the level of school and family, could have

Table 2. Answers to the Question ‘How Important a Role did Politics Play in Your Childhood Environment?’ by Generation (Percentage) Total Very or fairly important

Not very important

Not important at all

%

N

15

52

33

100

(168)

32

44

25

100

(124)

27

50

23

100

(274)

25

56

19

100

(129)

25

51

25

100

(695)

The generational of individual choice The generation of the suburban The generation of transformation The generation of war and reconstruction Total Note: χ2 = 18.38; d.f. = 6; p = 0.01.

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

9

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association Table 3. Answers to the Question ‘Have Your Parents Ever Offered Any Advice or Models about the Importance of Voting?’ by Generation (Percentage) Total

The generation of individual choice The generation of the suburban The generation of transformation The generation of war and reconstruction Total

Often

Rarely

Never

%

N

33 36 24 23 28

36 33 34 21 32

31 31 43 56 40

100 100 100 100 100

(154) (115) (254) (125) (648)

Note: χ2 = 26.69; d.f. = 6; p = 0.00.

meant avoidance of delicate political questions. This might have caused the fact that compared to the oldest generation, only a slightly larger percentage thinks that politics had a large role in their childhood. Advice from one’s parents about the importance of voting does not seem to be very common (Table 3). Some 40 percent of Finns have not received this kind of encouragement. However, it is worth noting that younger generations have received more advice than the older ones. This may be due to the fact that among the younger generations the first voting experience and pieces of advice related to it are still fresh in their mind. On the other hand, the lesser interest towards electoral participation among the young might have driven the parents to emphasize the importance of voting. The parents of current youth belong to a generation whose turnout is the highest among the generations – for example, in the parliamentary elections of 1999, turnout in the 45–54 age group was 75 percent and among 55–64 year olds it was 80 percent (Martikainen & Wass 2002, 47). As shown in Table 2, politics had the second largest role in the childhood and adolescence of those who are approximately 50 years old. Consequently, it appears that they are trying to pass on to their children the sense of duty to vote they themselves have adopted. Respondents were also asked whether they know which party their father and mother voted for during the respondents’ adolescence. Table 4 shows that 57 percent were able to name both their mother’s and father’s choice of a party, 13 percent knew either of them and 30 percent knew neither. The large portion of those in the youngest generation who are not aware of both of their parents’ party identification should, however, be studied further. The fact that 44 percent of the generation of individual choice is unable to name their parents’ party identification is a clear signal of the ‘privatization’ of politics. For instance, the eldest generation was born into a society divided by the civil war in 1917–1918. In that context, the parents’ party preference was reflected in the whole family (see, e.g. Jääsaari 1986, 263). Nowadays party choice seems to be a private matter, even inside the family, 10

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association Table 4. Knowledge of Mother’s and Father’s Choice of a Party during Respondents’ Adolescence by Generation (Percentage) Total

The generation of individual choice The generation of the suburban The generation of transformation The generation of war and reconstruction Total

Knows both

Knows one

Knows neither

%

N

44 56 64 65 57

12 19 13 12 13

44 26 23 23 30

100 100 100 100 100

(329) (200) (408) (195) (1132)

Note: χ2 = 52.99; d.f. = 6; p = 0.00.

and one’s political views are not necessarily reflected in other forms of action and cannot be seen from the outside. It is possible to give three alternative or supplementary interpretations for this phenomenon. First of all, it might be the case that it has become less common to talk about politics with family members. Second, even though some discussion takes place, children cannot ‘read’ or interpret their parents’ party choice based on their opinions due to the lack of civic literacy (see Milner 2002). The third option is that children simply are not interested in their parents’ political orientation anymore as they view their own political identity as being separate from their parents. In all of these cases, the inheritance of party identification (for Finland, see Grönlund et al. 2005a, 110–12) also becomes impossible because children cannot place their parents on the political map. Socialization into Electoral Participation during an Individual’s Life Cycle As noted earlier in this article, an individual’s political socialization does not have a certain end point, but instead continues throughout his/her life cycle (Alapuro 1969, 6; Jääsaari 1986, 172). Due to a change in an individuals’ life circumstances or societal change in general, attitudes towards electoral participation may be modified. It is possible to examine the differences in attitudes of persons of the same age during different periods of time via attitudinal statements such as ‘I think one fulfils his civic duty by voting regularly’ often stated in surveys. However, without a panel study, we cannot get a complete picture of how the views of a particular individual have changed during his or her life span. Table 5 examines the socialization into electoral participation during the individual’s life cycle. It needs to be pointed out, however, that we should be relatively cautious with this kind of retrospective variable as there might be some recall-call bias, especially where older people are concerned. Scandinavian Political Studies, Vol. 30 – No. 1, 2007

11

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association Table 5. Answers to the Question ‘How Much Has Your Attitude towards Participating in Elections Changed During Your Life?’ by Generation (Percentage) Total

The generation of individual choice The generation of the suburban The generation of transformation The generation of war and reconstruction Total

More positive

No change

More negative

%

N

44 42 25 31 34

51 42 56 53 52

5 15 20 16 15

100 100 100 100 100

(169) (125) (273) (131) (698)

Note: χ2 = 32.94; d.f. = 6; p = 0.00.

For instance, respondents might underestimate the amount of actual change. Consequently, the table shows that the attitude of every second Finn towards electoral participation has remained stable throughout his or her life. Moreover, a third of the respondents state that their attitude has moved into a more positive direction during their life course. Those whose views about electoral participation have changed in a more negative direction form only about 15 percent of the respondents. This observation is quite surprising given the decline in overall turnout during the past few decades. Interestingly, the proportion of those whose attitude has become more positive is larger among the two youngest generations than in the generation of war and reconstruction. Although this positive attitude among the younger generations does not necessarily seem be transformed into voting, as will be seen next, at the attitudinal level these cohorts are potential voters.

Effects of Generation and Socialization Factors on Turnout In order to form a more comprehensive picture of the differences in the socialization processes between different generations I will look at the effects of socialization factors on turnout. The main point is to find out how these factors affect the differences in turnout between various generations. First, Table 6 shows that turnout among the youngest generation (56 percent) is substantially lower than among the older ones, as expected. Moreover, clear differences in turnout can also be observed according to the role of politics during respondent’s adolescence, the knowledge of respondent’s mother’s and father’s choice of a party and the change in attitudes towards voting. Advice from parents concerning the importance of voting do not, however, seem to have a clear impact on turnout as the turnout among those who only rarely received this kind of encouragement is surprisingly high.5 12

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

Table 6. The Effects of Age and Socialization Factors on Turnout Multivariate models

Generation (gen) The generation of individual choice (18–33) 56 The generation of the suburban (34–43) 70 The generation of transformation (44–63) 78 The generation of war and reconstruction (64+) 75 Role of politics during adolescence (role) None 70 Relatively small 84 Large 87 Knowledge of mother’s and father’s choice of a party (know) Knows neither 54 Knows either 70 Knows both 79 Change in attitudes towards voting (change) More critical 53 No change 87 More positive 88 Encouragement to vote from one’s parents (encou) Not at all Rarely Often Constant N Nagelkerke R2 Cases predicted correctly (%)

86 97 96

Bivariate model

Model 1: sex + ses + adjusted

Model 2: sex + ses + gen + role

Model 3: + know

Model 4: + change

Model 5: + encou

1.00 1.78** 2.76*** 2.41***

1.00 1.62* 3.17*** 3.16***

1.00 1.23 2.48** 2.79*

1.00 1.19 2.05 2.28

1.00 2.05 3.46** 2.78

1.00 14.54* 8.14* 30.32**

1.00 2.28*** 2.81***

1.00 2.06** 2.88**

1.00 1.97** 2.61**

1.00 1.64 2.44*

1.00 1.33 1.76

1.00 1.78 0.67

1.00 2.00** 3.33***

1.00 2.33*** 3.14***

– – –

1.00 1.56 2.40**

1.00 1.39 2.00*

1.00 9.55* 1.46

1.00 6.20*** 6.67***

1.00 6.03*** 7.83***

– – –

1.00

8.28*** 20.97***

1.00 9.98** 26.66***

1.00 5.99*** 3.43**

1.00 6.27** 4.39**

– – – 0.48 662 0.15 82.7

– – – 0.05*** 578 0.35 86.6

1.00 1.37 1.26 0.40* 523 0.46 94.9

– – – – – – 0.29* 599 0.23 82.8

13

Notes: ***p < 0.001, **p < 0.01, *p < 0.05. The dependent variable is weighted separately for men and women in order to correspond with the official turnout from the electoral wards for the parliamentary elections of 2003. First, the bivariate models including generation and socialization variables are presented. The effect of these factors on turnout is shown in odds ratios. Each category of generation and socialization variables is compared to the reference group with an odds ratio of 1.00. Model 1 shows separately the odds ratios of voting in each category of generation and socialization variables after adjusting for sex and socioeconomic factors (ses). Socioeconomic factors include education (1 = primary education, lower secondary school; 2 = vocational school, upper secondary school, vocational college; 3 = higher college-level training, polytechnic degree; 4 = university, higher academic degree (MA)), social class (1 = manual workers, 2 = lower middle class, 3 = upper middle class, 4 = entrepreneurs, 5 = farmers, 6 = pensioners, 7 = students, 8 = unemployed, 9 = others) and income per household (from 1 = lowest quartile to 4 = highest quartile, quartiles not being forced to be exactly of equal size). Model 2 simultaneously includes the effects of sex, socioeconomic factors, generation and the role of politics during adolescence. In model 3 the knowledge of one’s parents’ choice of party is added. Model 4 includes the effect of attitudinal change towards voting and model 5 the encouragement to vote from one’s parents besides the aforementioned variables.

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

Turnout (%)

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

The bivariate models confirm these interpretations, as all of the differences between groups turn out to be statistically significant and follow the same pattern. In Model 1, sex and three socioeconomic variables (education, social class and income) have been adjusted. It shows a similar result to a previous study (see Martikainen et al. 2005): compared to the reference group of the youngest generation, clear differences between generations can be detected, even though the effect of generation is attenuated 21 percent among the generation of the suburban after adjusting for socioeconomic variables.6 The fact that the odds ratios for the generation of transformation and generation of war and reconstruction are larger than before the adjustment emphasizes the point stated earlier: the difference in turnout between the youngest and the eldest generations would be greater if there were no socioeconomic differences between them, given that the youngest generation is better educated (see Martikainen et al. 2005, 17). Model 2 shows that adjusting for generation and the role of politics during the adolescence, in addition to sex and socioeconomic factors, attenuates the differences in the odds ratios for voting between generations when compared to the youngest generation. However, the difference between the youngest generation and the generation of the suburban is not statistically significant. From the fact that the odds ratio for the generation of transformation attenuates 32 percent after adjusting the role of politics, and for the generation of war and want 17 percent, we can draw the conclusion that the difference in turnout between these generations and the youngest one can partly be explained by the lesser role of politics during childhood and adolescence of the generation of individual choice. In Model 3, the respondent’s knowledge of his or her parent’s choice of a party during the respondent’s adolescence has been included, but the results are not statistically significant. In Model 4, the change in attitudes towards voting has been included in the analysis. Compared to the previous model, this new variable increases the odds ratios for the generation of transformation (OR = 3.46). In order to interpret this finding, it should be noted that among the reference group, the generation of individual choice, 44 percent of respondents’ attitudes towards voting have changed into a more positive direction during their lives, as the corresponding figure is only 25 percent for the generation of transformation. It is quite interesting that adding this variable into analysis increases the original differences in odds ratios between the youngest generation and the generation of transformation as compared to Model 1. The result suggests that if the youngest generation had not faced this change in attitudes into a more positive direction, the difference would be even larger. In Model 5 we can see that adjusting for encouragement to vote from one’s parents, the differences in odds ratios between the youngest generation 14

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

and the other generations increases remarkably, especially when difference between the youngest and the oldest generation is considered. However, as stated earlier, we should be careful not give too much importance to this result due to problems with the variable concerned (see Note 5). Nevertheless, we can conclude that parental advice about the importance of voting has a similar impact as attitudinal change towards voting, as the differences in the likelihood of voting would be higher between the youngest and the preceding generations unless the youngest one had not got that much encouragement to vote.

Conclusions In this article I have examined socialization into electoral participation among four Finnish generations and studied the extent to which political socialization is able to account for the generational differences in turnout. The results show that the socialization processes of different generations clearly differ from one another. However, these differences are not straightforward and, consequently, interpreting them is relatively challenging. Moreover, the results show that generational differences in electoral participation cannot be explained by the differences in socialization, at least in the form it was operationalized in this study. When looking at early socialization, we can see that the role of politics in childhood and adolescence clearly has been the smallest in the adolescence of the youngest generation, the generation of individual choice. However, this generation has also received the most advice about the importance of voting from their parents. Most of their parents are in their 50s and belong to a generation in whose adolescence politics had a larger role and whose turnout is higher than the turnout at the aggregate level. Consequently, it seems that this generation of transformation has tried to pass on to their children this sense of obligation to participate. The knowledge of parents’ choice of a party is remarkably less frequent among the youngest generations than among the older ones, which is a clear indicator of the privatization of politics. All in all, excluding the advice to vote, the results show that the role of the family and childhood environment as socialization agents has declined with time. It is also worth noting that, at the level of attitudes, the youngest generation has more positive feelings towards electoral participation than the oldest one. Moreover, attitudinal change towards voting during an individual’s lifespan has been the most positive among the youngest generation. While the attitude of every fifth person in the generation of transformation towards electoral participation has changed into a more negative direction throughout their lifespan, the corresponding figure among the youngest generation is Scandinavian Political Studies, Vol. 30 – No. 1, 2007

15

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

only 5 percent. However, among every generation, those whose attitudes have remained the same form the majority. Finally, I examined how socialization-related factors affect the differences in turnout between generations. The results show that the differences in turnout between the generational of individual choice and the preceding generations would be substantially higher when taking into account only sex and socioeconomic factors. This is due to the facts that even though politics has had the smallest role during the formative years of the youngest generation and they most often do not know their parents’ partisanship, this generation is clearly better-off regarding parental advice about the importance of voting and change in attitude towards electoral participation. Nevertheless, their low turnout shows that they still cannot find enough stimulus to actually cast a vote. As has been shown earlier (Martikainen et al. 2005) and confirmed here, low turnout among the young cannot be explained by their lower socioeconomic status. Consequently, the explanation for the tendency of the young generation to vote less than the previous generation needs to be further examined. Two additional remarks have to be made. First, this study might be criticized for the choice of indicators concerning political socialization. For instance, Plutzer (2002) has found that the strength of parental partisanship has an enduring effect on children’s likelihood of voting. Building on a developmental theory of turnout, Plutzer finds that ‘if adults in their 30s, 40s or 50s become disenchanted with parties, they tend to continue voting if that was their habit. However, the results suggest that their children will be slightly less likely to vote (perhaps decades later) as a consequence.’ Unfortunately, the Finnish National Elections Study 2003 does not include such a variable. On the other hand, the operationalization of the concept used here is already very much concentrated on experiences during adolescence. In order to widen the analysis, it could be possible to examine the effect of political knowledge. Although political knowledge rises with age as a person becomes more interested in societal matters (Milner 2002, 47), there are also clear signals for a generational effect. Many studies in the United States have shown that the youth of today know considerably less than previous generations during their adolescence (Delli Carpini & Keeter 1996; Milner 2002, 47). As Putnam (2000, 36) notes, from the earliest opinion polls in the 1940s to the mid-1970s, the young were at least as well informed as their elders. Although political knowledge is a very intriguing subject and the data used here include variables that measure it, I decided to exclude it from the analysis. This was due to the fact that political knowledge is, along with age, the most important factor explaining turnout (Grönlund et al. 2005b, 137–8) and would have thus ‘overpowered’ the impact of other socialization-related variables. Consequently, the result probably would have been that political socialization perfectly explains the generational effect in turnout, which is not the case, as we have seen. 16

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

Second, political socialization is related here solely to electoral participation. It would be interesting to examine how some socialization-related variables are connected to other forms of political participation such as signing a petition, joining a boycott or participating in a demonstration. On the other hand, as the main purpose of this study was to examine the extent to which socialization accounts for generational differences, this approach would have been fruitless given that there is no evidence of generational effect in other forms of political participation based on longitudinal data, at least so far. However, in the future, this information would be particularly valuable for forming a comprehensive picture of the role of generational effects in political participation.

ACKNOWLEDGEMENTS This article is part of a research project ‘Elections and Representative Democracy in Finland’ funded by the Academy of Finland (project 8104411). I warmly thank the leader of the project, Heikki Paloheimo, Mikko Mattila, Semi Purhonen and the anonymous referees for their valuable comments on earlier versions of the article.

NOTES 1.

2.

3. 4.

5.

6.

The empirical modelling of these three time-related factors is somewhat problematical due to perfect multicollinearity (see, e.g. Toivonen 1999; Glenn 2005). However, it is beyond the scope of this article to discuss this problem and its various solutions in more depth. Delli Carpini, as many others, refers to German sociologist Karl Mannheim, whose essay on generations was first published in 1928 and translated into English in 1952. The gist of Mannheims’ analysis was to distinguish biological generations from those generations based on common, societal experiences (see, e.g. Purhonen 2005). I thank Kyösti Pehonen for this point of view. The interpretation of the key experiences of the suburban generation presented here is quite different from the original by Roos (1987). According to Roos, the lack of experience has been the common feature for the generation of transformation. However, it seems reasonable to assume that ease of life has been predominant during the adulthood of the older generation and, consequently, the key feature of adolescence for their children who form the generation of the suburban. The proportion of missing values in this variable was particularly large (N = 649) and, when cross-tabulated with the turnout variable, N was further reduced (to 589). Consequently, it is possible that this variable seriously skews the whole model. However, it was decided not to use missing values as a separate category and include them in the model. This is because some of the missing values are due to the fact that this question was asked in a supplementary, self-administered questionnaire that consists of only 59 percent of those interviewed face-to-face, and some of them are ‘true’ missing values. Because of this problem, the variable is not in its chronological place after the role of politics in Table 6, but instead in the final model 5. Percentage attenuation after adjusting for other variables can be calculated using the formula: 100 * ((ORi1 − ORi2)/(ORi1 − 1)), where ORi1 denotes the odds ratio for the variable obtained from the unadjusted model and ORi2 denotes the odds ratio for the same variable I obtained from the model that also includes the adjusted variables. I owe Pekka Martikainen for this calculation (see Martikainen et al. 2005).

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

17

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association

REFERENCES Abramson, P. R. & Aldrich, J. H. 1982. ‘The Decline of Electoral Participation in America’, American Political Science Review 76, 502–21. Alapuro, R. 1969. Sukupolvet ja arvot. Helsinki: Helsingin yliopiston sosiologian laitoksen tutkimuksia 118/1969. Blais, A., Gidengil, E., Nevitte, N. & Nadeau, R. 2004. ‘Where Does Turnout Decline Come From?’, European Journal of Political Research 43, 221–36. Borg, S. 2004. Mahdollisuuksien maa. Kartoitusta ja puheenvuoroja suomalaisen kansalaisvaikuttamisen tutkimuksesta. Helsinki: Oikeusministeriön julkaisu 2004:10. Brody, R. A. 1978. ‘The Puzzle of Political Participation in America’, in King A., ed., The New American Political System. Washington, DC: American Enterprise Institute. Campbell, A., Converse, P., Miller, W. & Stokes, D. 1960. The American Voter. New York: John Wiley & Sons. Connell, R. W. 1972. ‘Political Socialization in the American Family: The Evidence Re-examined’, Public Opinion Quarterly 36, 323–33. Dahl, R. 1961. ‘The Behavioral Approach in Political Science: Epitaph for a Monument to a Successful Protest’, American Political Science Review 55, 763–72. Dalton, R. J. 1988. Citizen Politics in Western Democracies: Public Opinion and Political Parties in the United States, Great Britain, West Germany, and France. Chatham, NJ: Chatham House. Delli Carpini, M. X. 1986. Stability and Change in American Politics: The Coming of Age of the Generation of the 1960s. New York/London: New York University Press. Delli Carpini, M. X. & Keeter, S. 1996. What Americans Know about Politics and Why It Matters. New Haven, CT: Yale University Press. Dudlay, R. L. & Gitelson A. R. 2002. ‘Political Literacy, Civic Education and Civic Engagement: A Return to Political Socialization?’, Applied Developmental Science 6, 175–82. Franklin, M. N. 2004. Voter Turnout and the Dynamics of Electoral Competition in Established Democracies since 1945. New York: Cambridge University Press. Franklin, M. N., Lyons, P. & Marsh, M. 2004. ‘The Generational Basis of Turnout Decline in Established Democracies’, Acta Politica 39, 115–51. Friedman, L. N., Gold, A. R. & Christie, R. 1972. ‘Dissecting the Generation Gap: Intergenerational and Intrafamilial Similarities and Differences, Public Opinion Quarterly 36, 334–46. Glenn, N. D. 2005. Cohort Analysis. Thousand Oaks, CA: Sage. Greenstein, F. I. 1970. ‘A Note on the Ambiguity of “Political Socialization”: Definitions, Criticisms and Strategies of Inquiry’, Journal of Politics 32, 969–78. Grönlund, K. et al. 2005a. ‘Kiinnittyminen politiikkaan’, in Paloheimo, H., ed., Vaalit ja edustuksellinen demokratia Suomessa. Helsinki: WSOY. Grönlund, K. et al. 2005b. ‘Äänestysosallistuminen’, in Paloheimo, H., ed., Vaalit ja edustuksellinen demokratia Suomessa. Helsinki: WSOY. Highton, B. & Wolfinger R. E. 2001. ‘The First Seven Years of Political Life Cycle’, American Journal of Political Science 45, 202–9. Hyman, H. 1959. Political Socialization. Glencoe, IL: Free Press. Inglehart, R. 1977. The Silent Revolution: Changing Values and Political Styles among Western Publics. Princeton, NJ: Princeton University Press. Jääsaari, J. 1986. ‘Sukupolvet, elämäntapa ja politiikka’, Politiikka 28, 258–81. Jääsaari, J. & Martikainen, T. 1991. Nuorten poliittiset valinnat. Helsinki: Gaudeamus. Ljungberg, E. 2003. ‘Politisk socialisation – ett forskningsområde som söker sin identitet’, Politiikka 45, 133–47. Lyons, W. & Alexander, R. 2000. ‘A Tale of Two Electorates: Generational Replacement and the Decline of Voting in Presidential Elections’, Journal of Politics 62, 1014–34. Martikainen, P., Martikainen, T. & Wass, H. 2005. ‘The Effect of Socioeconomic Factors on Voter Turnout in Finland: A Register Based Study of 2.9 Million Voters’, European Journal of Political Research 44, 645–69. Martikainen, T. & Wass, H. 2002. Äänettömät yhtiömiehet. Osallistuminen vuosien 1987 ja 1999 eduskuntavaaleihin (Vaalit 2002:1). Helsinki: Tilastokeskus. Martikainen, T. & Yrjönen, R. 1991. Vaalit, puolueet ja yhteiskunnan muutos (Tilastokeskus, Tutkimuksia 178). Helsinki: Tilastokeskus.

18

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

© 2007 The Author(s) Journal compilation © 2007 Nordic Political Science Association Milbrath, L. W. 1965. Political Participation: How and Why Do People Get Involved with Politics? Chicago, IL: Rand McNally. Milner, H. 2002. Civic Literacy: How Informed Citizens Make Democracy Work. Hanover, NH/ London: Tufts University, University Press of New England. Niemi, R. G. & Sobieszek, B. I. 1977. ‘Political Socialization’, Annual Review of Sociology 3, 209–33. Perea, E. 1999. ‘Youth Turnout in National Parliamentary Elections’, in Youth Voter Participation: Involving Today’s Young in Tomorrow’s Democracy. Stockholm: International Institute for Democracy and Electoral Assistance (IDEA). Pesonen, P., Sänkiaho, R. & Borg, S. 1993. Vaalikansan äänivalta. Tutkimus eduskuntavaaleista ja valitsijakunnasta Suomen poliittisessa järjestelmässä. Helsinki: WSOY. Plutzer, E. 2002. ‘Becoming a Habitual Voter: Inertia, Resources and Growth in Young Adulthood’, American Political Science Review 96, 41–56. Purhonen, S. 2002. ‘Sukupolvikäsitteen kolme ulottuvuutta. Diskursiivisen dimension merkitys sukupolvitietoisuuden rakentumisessa’, Sosiologia 39, 4–17. Purhonen, S. 2005. ‘Onko Attac sukupolviliike? (Ja mikä sukupolviliike oikein on?)’, in Hoikkala, T., Laine, S. & Laine, S., eds, Mitä on tehtävä? Nuorison kapinan teoriaa ja käytäntöä. Helsinki: Loki-kirjat. Putnam, R. D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Renshon, S. 1992. ‘Political Socialization’, in Hawkesworth, M. & Kogan, M., eds, Encyclopedia of Government and Politics, Vol. 1. London: Routledge. Roos, J.-P. 1987. Suomalainen elämä. Tutkimus tavallisten suomalaisten elämäkerroista. Helsinki: Suomalaisen Kirjallisuuden Seura. Rubenson, D. et al. 2004. ‘Accounting for the Age Gap in Turnout’, Acta Politica 39, 407–21. Sapiro, V. 2004. ‘Not Your Parents Political Socialization: Introduction for a New Generation’, Annual Review of Political Science 7, 1–23. Setälä, M. 2003. Demokratian arvo. Teoriat, käytännöt ja mahdollisuudet. Helsinki: Gaudeamus. Tingsten, H. 1937. Political Behaviour: Studies in Electoral Statistics. London: P. S. King. Toivonen, T. 1999. Empiirinen sosiaalistutkimus. Filosofia ja metodologia. Helsinki: WSOY. Van Deth, J. W. 2001. ‘Studying Political Participation: Towards a Theory of Everything?’ Paper presented at the ECPR Workshop on Electronic Democracy: Mobilisation, Organisation and Participation via new ICTs, Grenoble. Wass, H. 2006. ‘Life-cycle, Generational and Period Effects in the Parliamentary Elections of 1975–2003 in Finland’. Unpublished manuscript. Wilska, T. 2004. ‘Nuorten kollektiivinen sukupolvitietoisuus nuorisobarometrin valossa’, in Wilska, T., ed., Oman elämänsä yrittäjät? Nuorisobarometri 2004. Helsinki: Nuorisoasianneuvottelukunta, julkaisuja 28. Nuorisotutkimusverkosto/Nuorisotutkimusseura, julkaisuja 44.

Scandinavian Political Studies, Vol. 30 – No. 1, 2007

19