Conceptual and methodological issues about young people and politics

Transcription

1 Conceptual and methodological issues about young people and politics Irene Martín & Gema García-Albacete Departamento de Ciencia Política y Relaciones Internacionales Universidad Autónoma de Madrid Work in progress. Please, do not quote without the authors permission. ECPR Joint Sessions, Nicosia 2006 Workshop Nb. 5: Studying Forms of Participation

2 1- Youth and Politics: state of the art and research questions During the last years, several institutions and researchers have shown their concern about both the growing distance between citizens and the institutions and channels for democratic participation and their diminishing support for democratic institutions. Evidence has been found of decreasing levels of electoral participation, increasing critical attitudes towards politics and a growing gap between citizens and conventional forms of political participation (Blais, Gidengil, & Nevitte, 2004; Dalton, 2004; Dalton & Wattenberg, 2000; Pharr & Putnam, 2000). Beyond strictly political attitudes and participation, some authors have also pointed out to a decline in social capital (Pharr & Putnam, 2000; Putnam, 2000). This last statement has given way to an intense debate concerning whether it is really a decline in civic engagement we are facing or whether it is just a transformation of social and political values, attitudes and forms of participation. Neither can we find agreement amongst the different authors about the implications these changes can have on the quality of democracy (Stolle & Hooghe, 2005a) 1. The observed transformations in the relationship between citizens and politics seem to be especially marked amongst the youth. More specifically, attention has been paid to the low turnout rates among the new generations in a few western democracies (Blais et al., 2004; IDEA, 1999; Saha, Print, & Edwards, 2005). Given the consequences this phenomenon could have for the functioning of democratic institutions in the future, several works have seen the light lately that deal with the political attitudes, values and behaviours of the younger generation. These studies have been especially prolific in USA, the United Kingdom, Canada and France. When analysing the political characteristics of the younger generations, some or all of the following patters have been found: younger citizens are less trusting both of their fellow citizens and of their government, they are less interested in politics and public affairs, less informed about the political institutions and the political process, and less willing to participate in political arena (Bennett, 1997; Blais et al., 2004; Delli Carpini, 2000; Pirie & Worcester, 1998, 2000). The explanations pointed out for this phenomenon can be summarized in four main not necessarily self-exclusive - arguments. First, their higher level of cognitive mobilization. Secondly, the different attitudes and political participation patterns of young people could be explained by the life-cycle effect. The third one would stress the increasing disaffection of young people towards politics. Finally, a fourth argument refers to a wider cultural change beyond politics. On one hand, the lower propensity of young people to vote could be due to their higher level of education. In other terms, it could be due to their higher level of cognitive mobilization and their more critical stance towards electoral politics (Barnes, Kaase, & al, 1979; Inglehart, 1990; Verba, Nie, & Kim, 1980) Secondly, the idea that these differences are a consequence of the life-cycle effect is present in an explicit or implicit way in most of the literature. According to it, young citizens, who have not been incorporated to adult life in social, economic and work related terms yet, have dissimilar concerns and interests, politics wouldn t be one of those. 1 For a review of this debate see Stolle y Hooghe (2005) 2

3 From a different perspective, some researchers have called the attention to the great compromise of young citizens towards elections and the democratic process. However, this wide support towards the democratic ideal is compatible with increased distrust towards political parties as a mechanism for political participation (EUYOUPART, 2005). Some have even suggested that this lack of confidence is due to the failure of the political system in providing enough stimuli to involve the new generations in politics, not allowing them to play a larger role in the public sphere (Henn, Weinstein, & Forrest, 2005). The fourth group of explanations have addressed a broader cultural change. This change would be the result of social and economic transformations among others, increasing working hours, changes in family roles, technological development, globalization and geographical mobility that have driven citizens to a search for more individualistic and lifestyle coherent forms of political engagement (Bennett, 1998). This view would be in line with the idea of the shift towards postmaterialist values in post-industrial democracies (Inglehart, 1990). According to these theses we wouldn t be witnessing a decline in civic and political engagement but an adaptation of young citizens to new circumstances they happened to socialize and live in. This adjustment would mean a moved away from traditional forms of participation together with the rise of new forms of interest and political engagement. Given the transformations in the relationship between citizens and politics and its adjustment to new lifestyles and interests, variations in the meanings people give to politics and the appearance of new mechanisms used to influence politics seem plausible. In this context several authors have indicated the limitations of the literature of political participation in explaining the link between youth and politics, as it usually concentrates in studying the traditional forms of political participation (Gauthier, 2003; Kovacheva, 2005; O'Toole, Lister, Marsh, Jones, & McDonagh, 2003). In order to adapt research to the evidenced changes in values and to understand new generations political involvement, attention has been place on new forms of political participation. These emerging forms of political involvement would embrace political actions such as political consumerism buying or boycotting products because of ethical, environmental or political reasons -, Internet- promoted campaigns, and sending protest messages using other new technologies such as and sms. As the literature has stated, these emerging forms of political engagement differ from the conventional ones in several dimensions. Their structure is more flexible and egalitarian. They are also different as to the kind of implication required: participation is more individualistic, immediate and spontaneous. Actions usually concern environmental, human rights and social justice issues. Their targets are also different: activities are commonly directed beyond political institutions and government, addressing companies and international organizations. Finally, their scope frequently goes beyond the national sphere and involves more than one country (Stolle & Hooghe, 2005b) 2. All these characteristics would make the emerging forms more attractive to young adults and, specially, to those more disaffected with traditional channels (Micheletti & Stolle, 2005). Is youth political involvement via these emerging forms compatible with their participation in conventional forms? Or does this option imply turning down traditional forms? In the seventies the rise of protest political actions did not suppose the decline of 2 Recent studies on political consumerism are: (Micheletti, Follesdal, & Stolle, 2004; Stolle, Hooghe, & Micheletti, 2005; Stolle & Micheletti, 2005) 3

4 conventional participation (Barnes y Kaase, 1979). Can we say the same regarding nowadays youth? This question has been addressed by political science studies with different results. Norris (2003) has concluded that, effectively, young people tends to involve in caused-oriented activities 3 while moving away from electoral participation and implication in political parties. On the contrary, Stolle and Hooghe (2005b) have stated that, while in the seventies youth was the main protagonist in non-conventional participation activities, in 2002 they are only the more numerous group in demonstrations, but not in other emerging forms of political participation. What we intend to do in this paper is to reassess what are the trends in political participation amongst young Europeans in comparative terms. For this purpose, we will analyse the data from the two waves of the European Social Survey that have been carried out so far 4. Our paper is structured as follows. First, we will present some descriptive analysis in order to map the political involvement of different age groups in Europe, and we will answer some preliminary questions. Do youngsters participate more or less than older generations? Are there differences in youth political involvement depending on the kind of participation we analyze? Secondly, we will compare the influence of some sociodemographic and political factors on different types of political participation in several European countries. Our main intention is to see whether young people s electoral turnout is related to their position in the life-cycle, to their higher levels of political disaffection, or to their higher level of education. In this section we will also analyze whether the same factors can help us explain their participation in three other types of political actions: participation in demonstrations, boycott of certain products, and participation in social organizations. Finally we will pay attention to an alternative explication based on the political context. This section is still quite exploratory. In it we will examine the effect that the national and international political contexts can have on the political participation of different age groups. We explore the tentative hypothesis that there may be events that have a different qualitative and quantitative impact on the political participation of a specific age group. We conclude by summing up the main findings of our analysis and by suggesting some of the questions that remain open and that should be addressed in future research.. 2- Young people and different forms of political participation in 2004 As we know from previous research, the relationship between age and political participation is different depending on the type of participation we examine. While voting has a positive linear relationship with age (among others, Topf, 1995), forms of political protest present a linear negative relation. Other forms of political participation, such us political consume, contacts with politicians and social networks present a inverted U relationship with age (Ferrer, Medina, & Torcal, in press). Using a 3 Norris differentiates between citizen oriented actions voter turnout and participation related to political parties and cause oriented repertories political consumerism, signing petitions, demonstrations and illegal protests. 4 Unfortunately, there is no comparative available data specifically designed to study the relationship between youth and politics. In order to check the representativeness of the European Social Survey for the youth real population, the distribution of some socio-demographic characteristics have been compared to the official statistics for this group of the population in Spain. The small deviations contained in the sample makes us confident on the results of our analysis. 4

5 different political participation classification, Norris (2003) has pointed out that young people participate to a greater deal in cause-oriented actions but less than the older groups in citizen-oriented activities. For the purpose of our analyses we have chosen four different actions individuals can use to participate in politics 5 : having voted in the last general elections, having participated in lawful demonstrations, having boycotted certain products 6 and having worked in non-political organizations 7. Each of them is paradigmatic of a type of participation, namely, conventional participation, protest, new forms of participation and social capital. Our results corroborate that the relationship between political participation and age groups in European countries shows different patterns depending on the form of participation we focus upon (graphs 1-4). - Graphs 1 to 4 - Coherently with the literature, young citizens vote considerably less than previous generations in practically every country 8. The relationship between age and participation in demonstrations does not follow as uniform a pattern in different countries as the relationship between age and vote 9. However, we do observe a trend that is common to several countries, in the opposite direction of the previous one: the youngest participate to a greater extent in demonstrations than the other age groups 10.. Regarding the boycott of products, our data corroborate what the literature has already pointed out. Contrary to what theoretically could be expected for emerging forms of political involvement the most participative in boycotting products are the middle aged 11. Finally, participation consisting of working in non-political organizations 5 All of the questions belonged to the same battery and referred to whether the respondent had participated in these actions during the last 12 months prior to the survey. Therefore, they roughly refer to actions taken during the year Unfortunately, the European Social Survey doesn t include other emerging forms of political participation related to Internet campaigns or the use of other new information and communication technologies for political participation. 7 The indicator we use for having worked in non-political organizations is the less satisfactory of the four of them. The question referred to having worked in other organizations that were not included in the previous question (political party or action group). If we understand that the previous category included purely political organizations, then the one we have chosen would be likely to include the non-political ones. However, we are not sure whether action groups have been translated or understood as referring to political action groups in all countries. The 2 nd wave of the ESS does not include any other questions about belonging or participating in social organizations. 8 The chi square was significant in all countries except for The Netherlands. In all of the other countries the analysis of the corrected standardized residuals indicated that differences between the young and the rest were significant. In 10 of the 19 cases analyzed the relationship between voting and age followed a linear positive form. 9 The chi-square was not significant in 5 countries: Czech Republic, Hungary, Poland, Netherlands and Slovenia. In the United Kingdom, Greece, Norway and Sweden the residuals in the cells of the youngest group were not significant. 10 After analysing the residuals, we can say that there is a linear negative relationship between participating in demonstrations and age in six countries. 11 The relationship was significant in all countries but two. After analysing the residuals, several patterns became evident: in four countries (Austria, Germany, Switzerland and France) the relationship follows the shape of an inverted U and Finland is the only country where young people are the most participative in boycotting products. 5

6 and associations is not clearly related to age 12. However, when it is related, the youngest group appears to be less participative in non-political organizations than the older generations in several countries 13. Summing up, we find important differences in the levels of political participation of the younger generations depending of the form of involvement analysed. Younger citizens appear as the less likely to participate in elections and, in some countries, also in nonpolitical organizations. Also, they are definitely not the most participative when it comes to boycotting products. However, this age group is the most participative in lawful demonstrations. 3- Why is young people s participation different? Before presenting a multivariate model for political participation we have tested the bivariate relationship between some of the factors that explain political participation and age: Education, ideology, political interest, internal political efficacy and trust in political parties. As expected, education level is significantly higher among the younger group. The analysis of the aggregated levels of the left-right scale for all the countries shows that self placement of young people in the scale is not radically different in comparison to that of the middle generation. Anyway, it is worth noting that, compared to the older generations, the percentage of younger interviewees is higher in the centre category and lower in the centre right. Political interest is more clearly related to age in the direction we had expected, the younger generation declares being very or quite interested in politics significantly less than the two older groups while answering to be hardly interested or not at all interested with much more frequency that the middle age group. Regarding internal political efficacy and age, although not as strong as the one between political interest and age, we also found a significant relationship in the expected direction. Finally, and surprisingly enough, the relationship between trust in political parties and age is not as clear as the former ones. Trust in political parties is low for every age group and the pattern that this factor presents is similar for the younger and the middle age groups 14. As we have mentioned before, several explanations have been offered to explain why young people are less inclined to vote than the older age groups. We will put to test three of these: 1- Lower levels of turnout amongst the youth could simply be due to the fact that they are young. This means that today s young people will change their ways to participate in politics later during the life-cycle as they undertake family and job responsibilities that will contribute to their integration in the political community. 2- It could also be due to the fact that they have been socialized under different historical and political circumstances that have led them to adopt a more 12 The chi square was not significant for the following 7, out of 19 countries: Austria, United Kingdom, Hungary, Luxembourg, Poland, Portugal and Sweden. 13 This is the case in Finland, France, Ireland, Netherlands, Norway and Slovenia. 14 This is a surprising finding as we know for previous studies that there is a positive relationship between age and partisanship, and that the decrease of partisanship in advanced industrial democracies has been disproportionately concentrated among the young (Dalton & Wattenberg, 2000) 6

7 indifferent, distant attitude towards politics than the one observed in previous generations when they were young. Today s young people were born under democratic regimes and, therefore, they may give democracy for granted. Thus, their indifference. The political parties they know are not the old mass parties that had a distinct ideology and based their mobilizing strategies on a more direct contact with their supporters. Instead, the only parties they know are catch-all parties for which ideology and direct mobilization do not play such an important role anymore. If it is socialization under different circumstances that accounts for young people s lower electoral turnout, this trend will be distinctive of this generation for the rest of their lives. 3- Thirdly, the lower propensity of young people to vote could be due to their higher level of education. In other terms, it could be due to their higher level of cognitive mobilization and their more critical stance towards electoral politics. As we have already mentioned, some recent literature has shown a growing concern about young people s declining turnout. Some authors have suggested that young people are finding a replacement for conventional politics in social organizations - rising levels of voluntary work carried out by young people -, in actions of protest, and in new forms of participation. Our purpose now is to analyse, with the help of the most recent comparative data we have (the 2 nd wave of the European Social Survey), the following: a) Whether, in effect, young people s electoral turnout is related to their position in the life-cycle, to their higher levels of political disaffection, or to their higher level of education. b) Whether the same factors can help us explain their participation in actions that can be viewed as alternatives to party-politics or to electoral-politics such as participation in demonstrations, boycott of certain products, and participation in social organizations. For this purpose, we will carry out a multivariate analysis in which we will control the effect of belonging to a certain age group young (15-29), middle-aged (30-65) or elderly (+65)- for the effect of factors related to each of the three explanations just mentioned 15. In order to see if what hides behind the different levels of participation in the different age groups is the life-cycle effect we have included in our model several independent variables that describe situations that change during the life cycle such as whether one is living with his/her partner, has children at home, or has a paid job 16. But most explanations tend to emphasize the fact that today s young people are more politically disaffected than previous generations. To see if this is a plausible explanation of young people s political participation we have introduced the following three variables in our model: level of interest in politics, level of trust in politicians/political parties 17, and 15 The United Kingdom was excluded from the multivariate analysis since the education variable was not available. The interpretation of some of the coefficients is as follows: Gender (0=female;1=male); Education (ref. category: "Primary or less"); Age group (ref. category: "Middle aged (30-65)"); Ideology (ref. category (centre)). 16 The correlations between the first two variables are between 0,57 in Poland and Austria and 0,31 in United Kingdom. 17 In the 1 st wave of the ESS (2002) respondents were only asked about their level of trust in politicians. In the 2 nd wave (2004) they were asked about both their level of trust in politicians and in political parties. We have used the second but both indicators measure practically the same thing. The Pearson bivariate 7

8 the frequency with which one perceives politics as being too complicated to be able to understand what is going on (as a measure of internal political efficacy) 18. Additionally, we have also controlled for the effect of ideology. The fact that young people are the most participative age group in public lawful demonstrations and the one that votes the least could be due to the fact that they adopt more radical ideologies than older generations. But we are also interested in knowing whether a specific ideology left or right is related to each kind of participation. The possibility that it may be the different level of cognitive mobilization of the different age groups is also covered by introducing the level of education as an independent variable. Gender will also be controlled in our model. If, in effect, we find that young people s participation is totally explained by their position in the life cycle, the characteristics of their generation as we have defined them (political disaffection) and/or their higher level of education, then the relationship between age and political participation will disappear. In other words, it will not be statistically significant because it will be indirectly explained by the significant effect of these other variables. On the contrary, if after introducing all these controls, we find that the effect of age remains statistically significant and that the relationship is the same as the one described in graphs 1 to 4, then we will be quite sure that young people s peculiarities in what refers to political participation are not satisfactorily explained by the three above mentioned phenomena. Let us first focus on the model we have designed to explain participation in general elections thought the act of voting (table 1). - Table 1 - Our findings show a striking uniformity in practically all countries. After controlling for all the above factors, young people keep being less prone to vote than the older generations in all countries but one 19. This means that, independently of their level of education, of the fact that they still have no adult-like responsibilities, of whether they are interested in politics or not, and of their ideology there is some additional characteristic of this generation that keeps them from voting. The impact of education on the probability of voting is rather strong although maybe not as extended as we could have thought 20. This finding is relevant for our topic in the sense that it invalidates explanations that focus on the higher level of education of the youngest to explain their low levels of electoral turnout 21. Education, by itself, has a positive impact on voting. If young people are more educated then we would expect correlation coefficients in 2004 oscillate between 0,75 (Switzerland) and 0,89 (Greece, Spain and Czech Republic). 18 The correlations between these three variables have been analysed in order to avoid multicollinearity. There does not seem to be a problem in this sense. The highest bivariate correlation is the one between interest in politics and perceiving politics as complicated (values between 0,49 in Denmark and 0,27 in Greece). 19 The only exception is Sweden, were young people are not less likely to vote than the middle aged. 20 The impact of education is positive in just 11 of the 19 countries. 21 It also questions the idea that the level of education has barely any impact on electoral turnout (Topf, 1995). This idea is based on the fact that voting is a rather costless act and, therefore, resources such as education do not need to play a facilitating role as in other kinds of participation. 8

9 them to vote more, and not less, than older generations, at least in the countries were that positive influence exists. What about explanations based on the life-cycle effect? One of the factors that have to do with it living with a partner is related in a positive way with the probability of voting 22. However, the impact of this variable in isolation from the other two that measure the effect of the life cycle cannot be interpreted as unmistakable evidence of the impact of the life-cycle effect. It could also be interpreted as the positive role that social relations / networks can play in reducing the costs of participating. However, having a paid job, or having children living in the house both characteristics of the intermediate phase of the life-cycle do not have a positive impact on the probability of voting. Is the third explanation the one that blames the low levels of turnout of the youngest on a politically disaffected generation more plausible, then, than the other two? Yes, to a certain extent. Interest in politics has a positive influence on voting in almost every country 23. Taking into account that today s generation is more indifferent towards politics that previous one, this can explain why they participate less in elections than older groups. Finally, with regard to ideology, we do not find extended support for the idea that extremists are more inclined to vote. This happens in less than half of the countries and not always in both ideological extremes 24. Neither can we generalize about the link between a specific ideology and the propensity to vote. In five of the countries the probability of voting increases most amongst those that place themselves towards the left 25 and in seven of them it increases amongst those that place themselves in the right of the ideological scale 26. Also, we should be aware of the role the political context may play in activating the vote of a specific ideological sector. This implies that it will not necessarily be always the same ideological group the one most inclined to vote. We will come back to this issue in the last section of this paper. In sum, the fact that the variable that measures the relationship between being young and voting remains significant after controlling for all these other factor seems to support that sector of the arguments of literature that defines the peculiarities of the new generation, not in terms of their greater political disaffection, but in terms of other kind of values beyond the strictly political ones (Bennett, 1998; Inglehart, 1990). However, we should say as well that, according to our analysis and coherently with this literature this change of values not only affects the youngest sector of society. In several countries we find that there also remain significant differences between the oldest group and the middle aged: the oldest are more likely to vote than this last group 27. This would imply that the generational difference is not only present between the young and the rest but becomes already visible between those who were between 30 and 65, and those who were older than 65 in Let us now see which are the factors that influence the probability of participating in acts of protest such as lawful public demonstrations (table 2). 22 Living with a partner has a positive impact in all countries but three. 23 The only exception is Belgium, but this is probably due to the fact that voting is compulsory in this country. 24 Only in the Czech Republic we find that the probability to vote increases both in the extreme left and in the extreme right. 25 These countries were, in 2004, Austria, Spain, Hungary, Norway and Sweden. 26 In 2004 these countries were the Czech Republic, Finland, France, Ireland, Poland, Portugal and Slovenia. 27 This happens in 12 out of 19 countries. 9

10 - Table 2 - After controlling for the factors related to the three proposed explanations, we see that in several countries the relationship remains the same as the one shown in the descriptive graphs: young people are more likely to participate in lawful demonstrations than the older age groups 28. Neither their higher level of education, nor the life cycle or their greater level of political disaffection can completely explain this phenomenon. Explanations based on the higher level of education of young people seem even less plausible when explaining participation in acts of protest than when explaining participation in elections. In general, higher levels of education do not have a positive impact on the probability of participating in demonstrations 29. Therefore, the fact that young people participate to a greater extent in this kind of acts cannot be explained by their higher education. Life-cycle explanations that sustain that, as people grow older, given their different circumstances and responsibilities, they will go less and less to acts of protest, do not seem plausible either. Having a paid job or having children at home do not affect negatively the probability of going to demonstrations. Living with a partner only diminishes this probability in a few cases. Neither can explanations around the idea of a politically disaffected generation help us explain why young people participate more than other generations in demonstrations. In all countries but two interest in politics increases the probability of going to demonstrations. If young people are less interested in politics than previous generations, how can we explain their preference for these acts of protest? We find that in several countries the probability of participating in acts of protest is highest amongst those that place themselves in the extreme left 30. The relationship between ideology and participation is therefore more clear when it comes to participating in demonstrations than when it comes to voting. We now want to see which is the relationship between age and one of the so-called new forms of participation: boycotting certain products (table 3). Could it be that young people, once we control for other factors, are more participative in this kind of acts than the bivariate relationship leads us to think? - Table 3 - The finding that it is the middle aged that more often participate in boycotting products is somehow diffused after controlling for all the independent variables included in our model. In fact, this kind of political participation is not clearly related to age. All we 28 We also confirm that the relationship between age and participating in lawful demonstrations is not as direct as in the case of voting. Young people are more likely to go to lawful demonstrations in 11 of the 19 countries. In four of the countries were being young is not significant (Czech Republic, Hungary, Poland and Slovenia) the low levels of participation in lawful demonstration and, therefore, the small number of cases, could be an explanation. In this case, the differences are between the youngest and the rest: contrary to what we had expected from the bivariate analysis, once controlled for diverse factors, the eldest are only less likely to participate in demonstrations than the middle aged in just 6 countries. 29 Education has a positive impact on participating in demonstrations in just 5 countries. 30 This happens in 10 out of 19 countries. 10

11 can conclude is that in three countries (Germany, France and Spain) the inverted U shape of the relationship is confirmed and that young people are the most participative group in boycotts in just two Nordic countries: Finland and Norway. Could it be that the relationship between young people and boycotting certain products is hiding behind the impact of other variables? As in the case of voting, we find that in several countries boycotting products is positively related with high levels of education 31. Once more, this would have made us expect that the age group with the highest levels of education the youngest would be the most participative one in the descriptive graphs above. However, this was not what we saw. Variables related to the life cycle are not at all related with this new form of political participation. Once more, we find that attitudes towards politics are the ones that matter the most. In almost all of the countries the probability of using this kind of participation increases the more one is interested in politics 32. This is in consonance with the low participation of young people in this kind of acts. But participation in boycotts also increases the more one distrusts political parties 33. This factor was not related to voting or to participating in lawful demonstrations. Boycotts, therefore, appear as the real alternative or new way of political participation. This makes sense to the extent that we often find political parties involved in the celebration of political demonstrations. We know that young people are less interested in politics and that we didn t find any evidence of them being more distrustful of political parties. Therefore, these two factors can explain why the younger generation doesn t participate more on political consumerism than the middle aged generation. On the other hand, boycotting has something in common with participation in lawful demonstrations: it is chosen mainly by those that place themselves in the left of the ideological spectrum (both on the extreme and in the centre left 34 ). Finally, which is the relationship between age and participation in non-political associations? Is it true that young people are finding a replacement to party-related forms of participation in social organizations through, for example, voluntary work? - Table 4- Our findings show that there is no relationship between doing work in organisations other than political parties or action groups and age. In some countries, however, we see that the probability of having worked for non-political organizations increases with the level of education 35. But then, we cannot explain why work in these organizations is not higher amongst the youngest in the descriptive graphs. None of the factors related with the life-cycle are significant but, once more, the positive impact of interest in politics can be generalized to almost all countries 36. Again, as in the case of boycotting products, it is reasonable to think that the lack of interest in politics amongst the youngest is related to their lower participation in social organizations. 31 This happens in 12 out of 19 countries. 32 This is the case of all countries but Slovenia, as we pointed out regarding participation in demonstrations, the low participation in political consumerism and the small number of cases could be the explanation. 33 This happens in 13 out of 19 countries. 34 In 12 and 10 countries respectively. 35 This happens in 13 countries out of The only exception being Hungary. 11

12 The conclusions we have reached are summarized in the following table: Voting Lawful demonstrations Boycott of products Social organizations LEVEL of young people s participation (descriptive) Low High Tends to low Tends to low Life cycle YES (?) NO NO NO Political disaffection YES NO YES YES Higher education NO NO NO NO Generational effect (New values?) YES YES NO NO 4- The international political context and the Spanish case as a point of departure We have mentioned as possible explanations of young people s peculiarities when participating in politics both the life-cycle effect, a generational effect based on the younger cohort s high political disaffection and their higher level of education. We have also suggested a fourth explanation that should be considered in future analyses: a generational effect based on the presence of new non-political values in the youngest group. However, there could be a fifth factor intervening in this relationship: a period effect. In other words, when analysing the levels of political participation of different age groups in one specific moment in time in our case the data of the second wave of the ESS were collected between November 2004 and April 2005 we cannot exclude the possibility that these levels were affected by the political context dominant in each country at that time. Period effects are understood as those changes in citizens relation to politics caused by a political event or series of events that take place in a limited period of time. What distinguishes them from life-cycle and generational effects is that they produce effects in the short term. These effects can be of two kinds: they can produce a change that will disappear after some time, but they can also produce long lasting effects. Period effects are more likely to affect political behaviours that political attitudes, which are known for being more stable in time. Therefore, when referring to political participation, it is a must to take into account the political context in which the survey was carried out. Often, period effects are assumed to affect all citizens equal. This is something we want to question in our paper. There may be events that have a higher impact on the political participation of a specific age group but not of others. Or, in any case, even when political events affect all age groups, the degree of their impact may differ in each of them. In principle there is a multiplicity of political events that may have affected the levels of political participation during the months, or the years, prior to Also, these events will often be specific of each country. During the months prior to the survey we are analyzing there was an event that affected all countries in a quite uncommon way: the decision by the so-called Coalition to invade Iraq taken in February This event mobilized millions of citizens around the world in acts of protest against this decision. It 12

13 was also the cause for campaigns to boycott American products. These events have supposedly mobilized especially young people of leftist ideologies. In spite of the almost unprecedented fact that mobilizations around the world took place in a simultaneous way and against one same cause, there were also national factors that added peculiar characteristics to how these events were lived in each country. The most evident one was the position of each government towards supporting or not the war, and towards sending troops to Iraq or not. There is another political event at the national level that usually contributes to alterations in political participation: the celebration of elections. Our purpose in this section is to explore which was the impact that the political events that took place between 2002 and 2004 had on young people s political participation and if these were similar in the other two age groups. In order to control for some of the political events that, beyond the mobilizations against the war could have been responsible for alterations in people s political participation, we will compare the data between the 1 st and the 2 nd wave of the ESS in those countries that celebrated general parliamentary elections during the months prior to the survey. The countries we have selected are: Country Date of Parliamentary elections Greece Position of incumbent Officially against (PASOK) Position of winner In favour (ND) Spain In favour (PP) Against (PSOE) Luxembourg Against (PCS-DP) Against (PCS-DP) Slovenia Against (LDS) In favour (SDS in coalition with 3 others) ESS2 Fieldwork to to to to We think that the mobilization of the youngest sector of society in different kinds of political participation (vote, demonstrations, boycotts, social organizations) needs to be researched in more depth. However, because it has been the kind of participation that seems to worry most in the literature about youth and politics here we will concentrate on the mobilization of youth in electoral participation. We will depart from the knowledge we have about what happened with the youth s electoral turnout in Spain. The indicator that measures the wellness of fit of our vote model is quite low in several countries, amongst which we find Spain. This may be an indication that what explains the levels of participation at the individual level in the past general elections are none of the three explanations we have explored but the political context. The incumbent party the centre-right Partido Popular was in favour of the invasion of Iraq. The main opposition party the centre-left PSOE was openly against it. During the last 2004 general elections turnout rose to levels considerably higher than those in previous elections. This increase was due, to a big extent, to the mobilization of young voters and, especially, of those that placed themselves towards the left of the ideological spectrum. These had been the most reticent to vote in previous elections. Could this be explained by the international environment of anti-war mobilizations? Or 13

14 did it have to do with national issues such as the mobilizations against the Prestige issue, the terrorist attack of March 11, the government s reaction to the terrorist attack, mobilizations against the government s reaction in the day prior to the election? It could be that one of the differences between young voters and the rest is that they are more likely to be mobilized by the international political context than by the national one as they are said to be more concerned with global politics. If it was the international political context that mobilised young leftist Spanish voters, it would seem reasonable to expect that: a) In those cases were a leftist government was in favour of the invasion of Iraq, a decline in turnout amongst the youngest in the left has taken place. This would have been a demonstration of the disapproval of the government s policies by those who had voted for it previous elections. Of the four cases we have selected, this could be the case of Greece to the extent that, beyond its official position, it gave military support to the Coalition. b) In those cases were a rightist government was in favour of the invasion of Iraq, a rise in turnout amongst the youngest in the left has taken place. This group would have been mobilized as a reaction against the government policies and in favour of an anti-war alternative. This would be the case of Spain. c) In those cases were a leftist government was against the invasion of Iraq, a rise in turnout amongst the youngest in the right has taken place. This would be the case of Slovenia. d) In those cases were a rightist government was against the war in Iraq, a decline in turnout amongst young voters of the right is expected. This would be the case of Luxembourg. The way we will put these hypotheses to test is by analysing the probabilities of voting of the different age groups of different ideologies (graphs 5-12). It goes without saying that, given the small number of cases we have, the character of this section is purely exploratory. Do the analyses support these hypotheses at all? Starting with Greece, the findings do not support our hypothesis (graphs 5-6). The turnout probability predicted by our model is considerably high especially as declared in surveys for all age groups and all ideologies. It is slightly lower in the left but the difference with the right is not really significant. There are barely any changes between 2002 and Two things can explain this situation. The fact that PASOK s government was ambiguous and not overtly in favour towards the invasion of Iraq may have kept it away from an important lost of electoral support from the youngest sector of its prior voters. But there is another national specific factor that can account for this lack of observable changes in spite of the international political context. This is the fact that vote in Greece is compulsory (at least according to the law, although not to its implementation). This may lead to an even higher overestimation of the vote in the past elections when measured through surveys than in the rest of the countries. In sum, surveys are not as good an instrument to study the levels of turnout in countries were voting is compulsory. What about Spain? Was the vote of young people in the left mobilized by the government s anti-war position? From the probabilities predicted, it is absolutely clear that turnout increased due to the mobilization of young voters placed in the centre, the centre-left and the left of the ideological scale (graphs 7-8). Turnout in Spain is lower amongst voters of the centre-left in general (Barreiro, 2002) but we find that it is mainly 14

15 concentrated in the young voters and that it is in this group that the changes were observed. Although it is very likely that it was not only the international political context behind this movement and that there were other national issues that contributed to it, the mobilization of this sector of the population is coherent with our hypotheses about the effect of the international political context. In the case of Luxembourg (graphs 9-10) we observe in the probabilities graphs that youngest group turned out to vote to a much higher degree than in previous elections but that there were no differences between those in the left, in the centre or in the right. This movement is not consistent with our expectation that the young people in the right would be demobilized. However, we have no explanation for the general rise in turnout. This could probably be explained by national issues of which we have no knowledge at the moment. Finally, in the case of Slovenia, we had foreseen a mobilization of the young voters in the right as a reaction to the anti-war stance of the centre-left government. The data do not clearly support our hypothesis for this country. What we see is a general decline in turnout. This decline in the probability to vote is especially acute amongst the youngest, although slightly less prominent for those in the centre-left and the centre-right. The cases we have analyzed do not consistently support our idea that the international context, in combination with the national context will mobilize certain ideological groups amongst the youth. However, we still think that this is a promising line of research to follow. One of the reasons could be the small number of cases we have analyzed. Also, we admit the excessive simplification when defining the combination of international and national political context by just referring to the position of national governments towards the invasion of Iraq. A case to case study would need to be done in order to improve the operationalization of complex contexts such as the electoral ones. Even if international factors may be more effective in mobilizing the youngest voters, national factors surely play a definite role as well. Anyhow, as we have learned from the Spanish case, generalizations about a general decline in the levels of electoral turnout amongst the youngest sector of society should not be anticipated before carrying out country by country analyses in which the political context is given a fair role in explanations about young people s political participation 5- Conclusions, discussion and further research questions In this paper we have tried to study the relationship between youth and politics in Europe by analyzing the more recent cross-national data that are publicly available at the moment. In spite of the evident work in progress character of this research we want to highlight some of the tentative conclusions we have reached so far. Our results corroborate previous studies in several ways. The need to distinguish different kinds of political participation and the implausibility of speaking about general trends in political participation is evidenced once more, no matter the age group we focus on. Also, we have seen that not only the levels but also the influence of different demographic and political factors varies from one kind of participation to the other. The levels of youth political participation in each of these types when compared to those of older generations are also confirmed: young people vote less, participate more in demonstrations and are not the most participative in other forms of political activities 15