Political or Institutional Disaffection? Testing New Survey Indicators for the Emerging Political Involvement of Youth

Transcription

1 Political or Institutional Disaffection? Testing New Survey Indicators for the Emerging Political Involvement of Youth Roger Soler i Martí roger.soler@gmail.com Department of Political Science and Public Law Autonomous University of Barcelona and Catalan Youth Observatory (OCJ) Government of Catalonia Abstract It is broadly accepted that in the last decades young generations have played an important role in the transformation of political participation patterns. However, some aspects in the nature of this change remain unclear: Quantitative studies have shown a decrease in political support and institutional participation of young people while some other works stress the increasing levels of non-conventional participation. Does this changes means that young people is more or less involved in social and political affaires? In the framework of the international economic crisis, the thought that governments and national political institutions are loosing sovereignty reinforces this institutional disaffection but, in the same time, this situation has been a source of political implication and mobilization particularly for young people as seen in the protest events that have taken place in Traditional survey indicators of political involvement reflect attitudes concerning traditional political objects like parties, politicians or institutions. Such indicators do not capture new forms of political involvement that are oriented more directly to the causes without the traditional mediation of political organizations or institutions. The paper analyzes a regional representative survey carried out just after the squares mobilization in Catalonia and including both traditional and new indicators about political involvement. Paper presented at the 22 nd World Congress of Political Science (IPSA). July 8-12, Madrid.

2 Political or Institutional disaffection? Testing New Survey Indicators for the Emerging Political Involvement of Youth Roger Soler i Martí The relationship that young people have with politics can seem contradictory. On one hand, messages in the media and many studies on political behaviour attribute to young people attitudes of disaffection and apathy towards politics. In fact, several longitudinal studies that have identified a decline in voter turnout and in other forms of formal political participation attribute part or all of this decline to the youngest generations (Franklin, Lyons and Marsh, 2004; Putnam, 2002; Pirie and Worcestern, 1998). On the other hand, recent decades have also witnessed the emergence of other forms of participation of a less institutional nature, such as protests and political consumption, in which the youngest generations have also played a prominent role (Norris, 2004, Zukin et. al. 2006). Given these changes, we might expect that there would also be a change in people s subjective forms of political involvement, and especially those of the youngest people. The changes in participatory behaviour, in apparently opposite directions, cannot be explained solely by an increase or decrease in political involvement. If there had simply been a decline in interest and psychological commitment to public affairs, we would not be able to explain why there has also been an increase in extra-institutional participation. Indeed, if there had been a net increase in people s psychological involvement in politics, we wouldn t see a decrease in voting or other forms of institutional participation. The literature has pointed out several factors that partially and approximately explain these changes: the increase in cognitive resources among the youngest generations, a change of dominant values, the prolongation of transition periods for young people, the Internet, the state s loss of political centrality, etc. But despite the importance that political attitudes have had on studies on political behaviour since their beginnings (Campbell et. al., 1960), survey studies have hardly paid any attention to a possible change in the forms of subjective political involvement that may be behind this change in patterns of participation. This paper builds on the idea that this is due to the fact that the traditional indicators of political involvement attitudes included in surveys do not allow for the detection of an emerging form of involvement. These indicators reflect attitudes oriented towards traditional political objects such as parties, the political class and institutions. That means that they do not capture new forms of involvement that, just like emerging forms of participation, are more directly cause oriented without the mediation of traditional political organisations or institutions. 2

3 In order to confirm the existence of different forms of political involvement, a battery of questions was included in a survey, carried out in Catalonia in 2011, which allow for the identification of different types of subjective political involvement. The new indicators are contrasted with traditional ones. Finally, their ability to explain different types of activism will be also analysed. The paper is structured as follows: the first part, of a theoretical nature, shows how people s relationship with politics has changed, especially with regard to political attitudes. An initial attempt is also made to theoretically characterise the type of political involvement that could be behind the emergence of certain forms of activism. This first section ends by explaining the four hypotheses that guide the analysis. The second section details the methodological aspects. The third part provides the empirical analysis and the findings obtained. First it lays out the analysis that gives rise to an alternative way of measuring political involvement that can distinguish between different forms based on their orientation. These indicators are compared with the most common indicators in survey studies to see how they are able or unable to capture the different forms of involvement. Next, it checks these new indicators ability to explain different forms of activism, controlling through traditional indicators. Finally, it studies whether young people effectively have some different forms of involvement that explain their greater predisposition to participating in forms such as protests or political consumption. 1. Theoretical discussion Changes in people s relationship with politics It has been three decades since Barnes and Kaase (1979) identified an expansion in non-conventional participation in Western democracies. Subsequent studies have gone on to confirm this trend. In most Western countries, there has been a clear expansion of political protests not just in terms of the amount of demonstrations and actions but also in the diversification of the of the kinds of individuals that use political protests as a form of political expression and participation (Norris, 2002; Dalton, 2002). Spain is one of the countries where protests, and especially attendance in demonstrations, are most widespread (Ferrer and Fraile, 2007; Jimenez 2011). 3

4 Meanwhile, the more institutional forms of participation, particular to representative democracy, seem to show an opposite trend. The extensive literature on voting has detected a decrease in voter turnout (Wattenberg, 2000; Aarts and Wessels, 2005; Patterson, 2002). In many countries, this drop in voting is also observed in other forms of participation, such as membership in parties and trade unions and belonging to formal associations in general (Putnam, 2002; Wattenberg, 2002). Behind a change in the forms of political participation, there should be a change in subjective orientations towards politics. From the Michigan School contribution (Campbell et. al., 1960), the study of political behaviour has devoted much attention to the influence that the subjective sphere exercises over behaviour. At the root of this, we know that a change such as that observed in patterns of political participation in recent decades should be linked with change in people s values and attitudes. But this possible change in the subjective sphere has not been analysed nearly as much by survey studies as changes in behaviour have been. A significant exception is Inglehart and his colleagues contribution in detecting and explaining cultural change (Inglehart, 1977 and 1990; Inglehart and Welzel 2006). While subject to much criticism, the theory of the emergence of post-materialist values has surely been the most influential contribution to the study of changes in values in post-industrial societies. According to Inglehart, advanced societies are experiencing a transition from traditional and survival-centred (materialist) values to secular-rational values and values of self-expression (post-materialist values). Inglehart has also linked this change with the expansion of protests, since this form allows for richer forms of individual self-expression than voting or mere membership in a political group of party. This greater freedom and precision in expressing individual preferences fits better with the emerging post-materialist values (Inglehart, 1990; Inglehart and Catterberg, 2002). Another academic contribution towards explaining the change in patterns of participation that includes a subjective factor is the one proposed by Dalton (2007 and 2008) that detects a change in the concept of what it means to be a good citizen in North American society. Dalton sees a shift from a concept of citizenship based on duty to one based on engagement, which is a different subjective experience of people s relationship with the public sphere, with an impact on their forms of participation. This explains why the notable decrease in voter turnout in the United States is not indicative of apathy and alienation from politics but rather of the diversification of the forms of participation in favour of more direct participation. 4

5 A common element in these changes is the central role of young people. Studies that stress the decline of traditional forms of participation point to the youth as one of the groups least disposed to political participation and most distant from institutions and parties (Pirie and Worcestern, 1998; Putnam 2002; Franklin, Lyons and Mash, 2004). However, authors that have given thought to the reinvention of political activism and the emergence of political protest have also highlighted the central role of young people (Norris 2004; Zukin et. al., 2006; Dalton, 2008). Finally, both Inglehart and Dalton see change in subjective orientations as based on the fact that young people have adopted different values and forms of dealing with politics, and that these forms have gradually become more present in society as a whole through a process of generational replacement. Change in political attitudes Changes in the forms of participation as well as those described by Inglehart and Dalton in political and civic values always show a twofold process going on. First, there has been a drop in traditional forms of relating to politics, while second and simultaneously, other ways of relating to politics are emerging. But there is a gap between values and behaviour where survey studies have not detected this twofold process: attitudes of political involvement. Survey studies on political attitudes tend to use indicators of attitude as independent variables of political behaviour. This means that, despite being subject to study for more than 50 years, analysis of the evolution of attitudes of political involvement has not received the attention that participation has. An exception to this lack of attention to the longitudinal study of attitudes of political involvement has been analysis of party identification. Indeed, identification with political parties has undergone a substantial change. Party dealignment, the decrease in people that say they feel they identify with a specific political party, is affecting most Western democracies (Fiorina, 2002, Clarke and Stewart, 1998). There may be many causes of this weakening of party identification, from the accumulated and increasingly public exposure of cases of corruption, the growing ideological ambiguity of large parties, the emergence of new political stakeholders, etc. According to Dalton (2002: ), there is another, more important factor that explains the consistency of this process throughout the years and its presence in different countries: what he calls cognitive mobilisation. This means that increasing levels of 5

6 education and access to information gives people better abilities to seek out, select and process political information by themselves without the need for political parties to mediate. With the exception of party identification, the literature has not given much attention to the evolution of subjective attitudes of political involvement. Still, based on the evolution of patterns of political participation and studies about values, one could form an idea of what the changes in people s subjective political involvement in Western democracies have been. Just as forms of participation particular to representative democracy and the traditional concept of citizenship as a duty have suffered a decline, particularly among the young, it may be expected that traditional forms of political involvement would do the same. Data on the drop in party identification would seem to confirm this hypothesis. In contrast, the increase in the forms of participation identified as non-conventional by Barnes and Kaase (1979), the emergence of post-materialist values and the appearance of a commitment-based concept of citizenship, especially among the young, should be linked with the appearance of an alternate form of subjective political involvement. Survey studies have not recorded or identified any indicator of attitude pointing in this direction. The thesis of this paper is that this circumstance is due to the fact that traditional indicators of political involvement (interest in politics, indicators of efficacy and party identification) refer to attitudes of sympathy and involvement with respect to the political objects particular to representative democracy (parties, institutions, elections, etc.) with a strong institutional component. Emerging forms of participation are more directly cause oriented without the mediation of representational institutions or organisations. That is why these classic indicators of political attitudes are not useful for detecting new forms of subjective political involvement. Another indicator of the problems encountered with traditional survey indicators when identifying emerging political involvement can be found in the results of some qualitative studies on young people s relationship with politics. In Great Britain, both the tradition of youth studies and social concern about a generational divide have stimulated the study of young people s relationship with institutions and the public sphere. Combining a panel survey study with qualitative techniques, Henn, Weinstein and Wring (2002) determine that the youth are interested in political problems and are not apathetic about politics in general. However, they do display a certain degree of scepticism towards formal politics. Alongside, O Toole and her colleagues (O Toole, 2003; O Toole et. al. 2004) expand on the meaning that young people attribute to politics and identify interests in more everyday and less 6

7 institutionalised aspects of politics. After conducting a qualitative study in Australia, Manning (2010) finds that youth feel distant from what they themselves identify as formal politics and do not perceive it as something related with their social and political commitments. In Catalonia as well, the youth especially seem to reject party politics (Gonzàlez, 2007). The shortcomings of classic indicators of political involvement in capturing new forms of involvement may be leading us to overstate young people s disaffection and apathy with respect to politics. Emerging political involvement The aim of this paper is to use survey questions to identify the existence of a subjective form of political involvement different from the traditional involvement captured by classic indicators. It must be characterised to do so. This form of political involvement should help to explain the emergence of forms of participation such as protests and political consumption. Thus, the elements used to explain the expansion of this type of participation should also be used to explain this form of involvement. Qualitative studies on youth and politics may also be a good source for deducing its characteristics. Taking these references into account, two great social changes can be identified that would explain the origins of this form of political involvement and may help to identify its characteristics. The first has a bottom-up dynamic; it arises among the citizens and has an impact on the political system. It refers to the change in individuals cognitive resources. What Dalton (2002: ) calls cognitive mobilisation and uses to explain the decline in party involvement probably has some broader effects in individuals relationship with politics. In recent decades, people s level of education has continued to rise, specifically regarding the youngest generations. We know that education has a positive impact on political participation because it provides individuals with resources and the civic awareness to understand (and thus become interested in) the world of politics, in addition to giving them greater capacity to influence it (Verba et al., 2002). These improved levels of education go hand-in-hand with a genuine revolution in access to information. Contrary to the widespread feeling that the media contribute to political confusion and apathy, some authors argue that the media play an important role by directly linking individuals with politics and that we have the best politically informed populations in history (Norris, 2000). The best exponent of this is the 7

8 Internet, which offers enormously diverse and rich ways to access political information. Aside from the debate over whether the amount of information favours or distorts knowledge, it seems clear that there is no longer just one or even a few different sources of political information; people can now choose the content and ideological perspectives of their political information from a very long catalogue. The second great transformation that shapes the forms of political involvement arises in the political system and conditions people s orientations (a top-down process). In recent decades, there have been substantial changes in the political systems of Western democracies that have had an impact on the distribution of political power, which is no longer monopolised by the state and its institutions. Crouch (2004) speaks of a post-democracy, where the state has lost centrality and the ability to make some of the most important political decisions in favour of international organisations, trans-national companies and financial interests. Management of the debt crisis that is especially affecting European countries is revealing that important political decision-making takes place outside of state institutions. The centre of political power is dissipating and is more difficult to identify. People are noticing these changes in the distribution of power and the political objects to which they give their attention are becoming more diverse. As such, there is an opening of the political object, which, while accelerating, is not a new phenomenon. The first studies on political behavior circumscribed participation to the electoral process. But in the early 1970s, Verba and Nie (1972) proposed a definition of political participation going beyond voting and elections. From then on, empirical studies have recognised increasingly diverse forms of participation, even coming to include political consumption, for example (Michelletti et. al., 2003; Stolle, Hooghe and Micheletti, 2005). The purchase or boycott of products is no longer just an action aimed at influencing the political elite that govern institutions. The political objects at which actions of participation are aimed are diversifying. Henrik Bang (2004) suggests that the political sphere is expanding, breaking down the boundaries between the political sphere and society in such a way that political orientations and expressions are manifested through people s daily lives. This means that the issues of a political nature that concern people are also diversifying and do not respond solely to traditional political cleavages (whether socio-economic, territorial or religious). Proof of this is the growing importance that some people give to new issues like those related to the environment or North-South inequalities. Reviewing the impact of these two great changes on the relationship between people and the political sphere, emerging subjective political involvement should be characterised by 8

9 two aspects: (1) it is directly oriented to causes or problems of a political nature and not to the political mediators particular to representative democracy (such as political parties and institutions); and (2) it is aimed at a new sphere of politics, at political topics that do not respond solely to traditional cleavages (whether socio-economic, territorial or religious). Hypothesis The main goal of this paper is to demonstrate that traditional indicators of interest in politics do not sufficiently capture the political involvement that leads to emerging forms of participation, namely those mostly used by young people. Four hypotheses have been developed to do so: (Hypothesis 1) Different types of subjective political involvement may be identified. Specifically, there is a type of involvement characterised by the fact that it is not aimed at traditional political conflicts and not aimed at institutional political stakeholders. (Hypothesis 2) Classical indicators of political involvement covered by surveys capture the different types of involvement unequally. They are better at capturing institutional political involvement and worse at capturing emerging political involvement. (Hypothesis 3) This form of political involvement is apt for explaining forms of participating in the emerging political activism (like protest and consumenrism) that is not captured by classical survey attitudes indicators. (Hypothesis 4) The explanatory shortcomings of traditional indicators of political involvement regarding political protest and consumption affect young people more than adults, since young people tend to use these forms of activism more. 2. Data and methodology The data used for analysis come from the Survey on Participation and Politics in Catalonia (EPP 11) (N=2.100). The EPP 11 is a telephone-based survey promoted by the Government of Catalonia s Ministry of Youth and Directorate General for Civic Participation and carried out by the Centre for Opinion Studies (CEO). It is specifically aimed at analysing the differences and similarities in the political participation and attitudes of young people and 9

10 adults. The fieldwork was conducted between July and September 2011 by surveying people more than 15 years old. The questionnaire contains an important repertoire of questions related to political attitudes and repertoires of participation. With regard to attitudes of subjective political involvement, it includes classic questions on the most common indicators of interest in politics, political efficacy, political discussion, party identification, etc. In order to identify different ways to deal with and get involved in politics, the questionnaire included a question on the interest that a battery of issues of a political and institutional nature aroused in the respondents. This was done to check if a distinction could be made between different types of political involvement based on interests and sympathy for different objects and topics of a political nature. The question was as follows: Below I will read you a series of topics and you should declare the extent to which you think they personally interest you on a scale of 0 to 10, with 0 meaning minimal interest and 10 meaning maximum interest. To what extent are you interested in... a. Immigration b. Elections c. North-South inequalities d. Political parties e. The rights of homosexuals f. The relationship between Catalonia and Spain g. The economy h. Security i. The environment j. Social policies k. The European Union l. Housing m. Access to employment Thus, this was an inductive strategy for identifying different types of sympathies and did not explicitly ask about any pre-established indicators of alternative involvement. The battery includes different political objects and topics with regard to which survey respondents could show different degrees of interest. These topics and objects includes elements that at first seem more suited to classic political involvement and others that may be more linked with the emerging political involvement that we are trying to detect. The analysis consists of identifying whether or not there are different types of political involvement behind the interests in these political topics and objects (hypothesis 1) and to characterise them. To perform analysis on hypotheses 2, 3 and 4, variables that the same 10

11 survey provides for traditional indicators of attitudes of political involvement and different forms of activism will be used. 3. Analysis and results Direct and institutional political involvement The first part of the analysis consists of identifying different forms of subjective political involvement. To do so, the EPP questionnaire had a question with a battery of political objects and topics in which survey respondents had to rate their level of interest on a scale of 0 to 10. According to the theoretical approach, the fundamental characteristic that should differentiate the various types of improvement is related to the objects and topics at which this involvement is aimed. As such, we need to see if there is a differentiated pattern in the interests shown by people; if there is some correlation between the rating given to each item proposed for the question that might denote different kinds of underlying political involvement. Applying a factor-based analysis gives rise to three components that explain most of the variance observed in the responses to the different items. The distribution of the weight of each issue in the components must express different ways of dealing subjectively with politics. [Table 1 about here] The interests are consistently distributed into three components: Component 1 explains a more important part of the variance and the issues with most weight are interest in North-South inequalities, the rights of homosexuals, immigration, social policies and the environment. In this component, the mediating stakeholders and institutions of representative democracy have very little weight. The most significant variables in the component are essentially interests aimed at emerging topics on the political agenda. Proposed topics also include those least related to traditional political cleavages. As such, this component captures emerging political involvement or direct political involvement. Component 2 deals with the variables of interest in political parties, elections, the European Union and the relationship between Catalonia and Spain, as well as interest in social policies, though to a lesser extent than component 1. This 11

12 component also features the items on the mediating stakeholders of representative democracy like parties, elections and an institutional entity like the European Union. The relationship between Catalonia and Spain and social policies that respond to territorial political and socio-economic cleavages, respectively, also appear to a lesser degree. Thus, it seems that factor-based analysis in this second component has captured political involvement aimed more at the institutional political system; institutional political involvement. Component 3 explains a smaller part of the variance and encompasses highly generic topics of interest such as security, the economy, housing, and access to information. These are very basic topics of interest to which all survey respondents usually give a high rating, so these are also the items that have the highest average on the scale of 0 to 10 and typically a lower deviation. As such, it seems that by grouping these basic issues together and distinguishing them from the others, this component captures a very basic and relatively undeveloped approach to politics; basic involvement. One way to better understand the significance of these factor-based components and put them to the test as indicators of political involvement is to analyse their relationship with the indicators that the survey studies usually use to measure attitudes of political involvement. To do so, a variable was created for each of the three components through factor-based ratings. The ratings were calculated with the Anderson-Rubin method, which generates a new variable with an average of 0 and a standard deviation of 1. [Table 2 and 3 about here] The results of the relationship between the new indicators of involvement and political attitudes confirm the interpretation of the components coming out of factor-based analysis. Right away the third component, basic involvement, has a negative relationship with all the attitudes of political involvement except party identification. This confirms the idea that this indicator covers a quite unsophisticated kind of approach to the political sphere; it is practically a counter-indicator of political involvement. However, institutional involvement shows a very strong positive relationship with the traditional indicators of political attitudes under study. For the third factor, direct involvement, the relationship is weaker and in some cases even negative. 12

13 These data also respond to the second hypothesis raised, according to which the traditional survey indicators of attitudes of political involvement capture the different types of political involvement unequally. Specifically, they do a better job of capturing the political involvement aimed at the political objects particular to representative democracy. Following this hypothesis, the indicator of institutional political involvement should have a strong positive correlation with interest in politics. In the other indicators, the survey respondents responding positively to the attitudes under study should show averages of institutional involvement higher than the mean (which is 0). However, the indicator of direct involvement should show a weaker relationship with the traditional indicators. Indeed, tables 2 and 3 indicate that the political attitudes traditionally included in surveys do a better job of capturing the new indicator of institutional involvement than direct involvement, particularly attitudes that most clearly refer to institutional political objects like external political efficacy (whose formulation refers explicitly to politicians) or party identity. In these same indicators, the average level of direct involvement is negative or neutral. In the rest of the attitudes, the average level of institutional involvement is still higher than that of direct involvement, with the sole exception of political discussion. Therefore, it seems that regarding the indicators of involvement generated through factor-based analysis, the questions on political attitudes often included in surveys do a better job of capturing institutional involvement than direct involvement. It also seems clear that the basic involvement component contains positions of indifference and apathy towards politics. Analysing the way the new indicators relate to the traditional ones, we can affirm that they capture one type of involvement better than others. Likewise, we need to see how these new indicators operate when explaining political participation to be able to check whether or not explanatory models of participation are missing the addition of a subjective kind of political involvement different than the one included in typical survey questions. The relationship between involvement and types of activism The main argument of this paper builds on the idea that different forms of political participation and activism respond to different forms of political involvement, and that emerging forms of participation, such as protest or political consumption, respond to a type of political involvement different from the involvement traditionally captured by the indicators commonly found in surveys. To confirm this argument, we have to analyse how the new 13

14 proposed indicators function with different forms of activism, and especially the ability of direct involvement to explain the emerging forms of participation. The first step consists of identifying the different forms of political activism. From the distinction made by Barnes and Kaase (1979) between conventional and non-conventional participation, the criteria for distinguishing the types of participation have changed, but what is found in one category or another seems to be the same. Barnes and Kaase distinguish participation based on whether they respected or transgressed the social norms and dominant values of society. Voting, party-linked activity and political contact are actions provided for and even promoted by the political system. However, other activities such as boycotting, squatting and demonstrations escape the mechanisms provided for in representative democracy and may involve a transgression of social norms. The distinction between what is conventional or not can be rather imprecise since what is considered transgressive regarding dominant values varies throughout time; it is historically relative. In fact, actions like demonstrations and petition signing have become so normalised that it doesn t seem they can be catalogued as non-conventional. Thus, the most recent attempts to classify forms of participation have sought out other criteria of distinction. Nevertheless, the final classification doesn t vary much. Torcal, Montero and Teorell (2006) base their distinction on the logic of representation, and Norris (2004) on their orientation. To classify the forms of participation on an empirical basis, a factor-based analysis was used on eight actions of political participation. Two components have come from this, the first of which deals with participation in demonstrations, petition signing, political consumption (the purchase and boycott of products) and expressive participation (wearing political messages on clothing, stickers, etc.). The second component includes party activity, political contact and voting. Adapting Norris terms (2004), the result distinguishes between institutional-oriented actions and cause-oriented actions. A variable was created for each type of activism. Institutional-oriented activism includes the survey respondents responding that they have contacted a politician or collaborated in a political party s activity. Despite the fact that voting also appears linked with these variables in factor-based analysis, it is not used because it is done at a very high frequency and does not help to distinguish among the activist profiles. To generate a variable for cause-oriented activism, there are more actions of even higher frequencies. To obtain variables with similar frequencies, the new variable for cause-oriented activism includes those individuals that stated they carried out three or more cause-oriented actions in the last year. 14

15 The dependent variable used in the analysis is created by cross-referencing both variables, resulting in the four categories that appear in table 4. [Table 4 about here] The table shows how both forms of activism are not exclusive. There are individuals who use institutional-oriented forms of participation just as much as cause-oriented forms of participation. This variable is used with four categories to see how the different types of involvement have an effect on activism. Thus, the people that only use cause-oriented forms are identified and they are expected to be those that show a more direct type of involvement. In the other hand, institutional-oriented activists should be influenced more by institutional involvement and by traditional political attitudes. Table 5 gives the average rating of the involvement indicators for each type of activism. [Table 5 about here] The results in table 5 follow in the third hypothesis. People who choose to use forms of participation such as protest or political consumption show levels of direct involvement clearly above the average and levels of institutional involvement slightly below the mean. This is exactly the opposite of what happens with people that choose institutional-oriented activism. Also of note is the high level of direct involvement of multi-activist people (those who use both institutional-oriented forms and cause-oriented forms). However, hypothesis 3 does not just predict that people who choose cause-oriented forms of participation have a direct type of involvement. It also states that common survey indicators do not capture this kind of involvement. As such, the models that explain political activism through political attitudes can be improved by including an indicator that covers direct involvement. Table 6 shows the results of two multinomial logistic regression models where the dependent variable is the typology of activism with the non-activist category as a reference. The first model uses political involvement indicators commonly found in most surveys on political participation and attitudes. The second model adds three new variables of involvement created from factor-based analysis. In the first model, the results largely respond to what one could expect, especially with regard to the institutional-oriented activism category: traditional attitudes of political involvement have a positive relationship with activism, with the exception of satisfaction with how democracy works, which seems to work the other way: being unhappy with how the system functions is an incentive to participate politically in any form. Yet in the first model, 15

16 the relationship between attitudes and cause-oriented activism appears less straightforward. The relationship is positive regarding interest in politics (while less for institutional-oriented activists) and with political discussion. On the other hand, the relationship to external political efficacy is negative and there is no significant influence with party identification. These two variables are those with a more explicit formulation aimed at institutional political objects. In the case of external political efficacy, the survey question used referred to politicians, while party identification obviously refers to political parties. Regarding political discussion and interest in politics, however, interpretation of the political object remained in the hands of the survey respondent, as the question referred only generically to politics. [Table 6 about here] Overall, the introduction of new variables of involvement in the second model improves the explanatory ability of the variance, as the pseudo-r 2 coefficient shows. The direct political involvement variable has no effect on institutional-oriented activism but does have a clearly positive effect on both cause-oriented activism and on multi-activism. It really seems that this form of involvement helps to explain the tendency to participate in emerging forms like protest and political consumption. In model 2 we can also see how the effect of the rest of the attitudes almost does not vary with respect to model 1. In other words, the introduction of the direct involvement variable does not cancel the effect of the traditional indicators and vice-versa; as such, they are explaining different attitudes. However, we cannot say the same regarding the institutional involvement variable. The effect of this variable is satisfactorily contained by classic variables of political attitudes. The only negative effect of this type of involvement appears in cause-oriented activism. Finally, the third component of factor-based analysis, basic involvement, shows a negative effect in all types of activism, which confirms its nature as a counter-indicator of involvement. The data confirm that there is a form of political involvement that has a clear effect on cause-oriented forms of participation like protest and political consumption. Moreover, a multi-variable analysis indicates that this form of political involvement is different than that covered by classic survey indicators. The youth and direct political involvement Literature on the drop in voter turnout and other forms of participation particular to representative democracy and literature that analyses the emergence of protest and political 16

17 consumption both give young people a leading role in the change. Studies on transformations in political orientation also attribute the change to the generation gap. As such, it seems that the youngest generations have a certain responsibility in transforming citizens relationship with the political sphere. By the dependent variable used, table 7 corroborates this greater tendency among the youth to use cause-oriented forms of participation. [Table 7 about here] The last of the hypotheses raised sustains that the youth s greater propensity to use cause-oriented activism is due in part to the fact that young people are more sensitive than adults to direct types of political involvement. This means that the young generations tend to have a psychological approach to politics less mediated by traditional political institutions and stakeholders and more directly aimed at emerging causes and themes. This makes them choose cause-oriented forms of participation, which is related with direct political involvement, as we have seen. To find out exactly what role direct political involvement plays for the youth, we repeat the regression model for types of activism with political attitudes as independent variables (with the new direct political involvement variable), introducing the youth as a dichotomous variable. This analysis excludes (basically young) individuals who have not left home for two reasons. First, the act of living with one s parents has a strong influence on political consumption. Young people still living at home are often less able to decide on the products that are purchased or no longer purchased. That affects the dependent variable component. The second argument for excluding individuals living with their parents is related with the generational effect of the change in political forms of involvement. The appearance of direct political involvement is part of the same process as the change in patterns of participation and in political values that the literature has basically attributed to the generation gap. The survey question used to generate the variables of involvement is new, so there are no longitudinal data to be able to separate the effect of the generational cohorts from that of the life cycle and the period (Glenn, 2005). By excluding those still living with their parents from the analysis, various phenomena related with youth transitions and linked with the life cycle vanish, which have a greater effect on attitudes and political behaviour (Stoker and Jennings, 1995; Lesthaeghe and Moors, 2000). In this way, the effect that the youth variable could have linked with the life cycle is minimised and, as a consequence but with caution, the effects can be linked more with the generational component. 17

18 [Table 8 about here] The results in the first model are very similar to those obtained so far for the different variables of political attitudes. It is noted that being young has an opposite effect for institutional-oriented activism than for cause-oriented activism. In the case of institutionaloriented activism, the coefficient is negative, indicating lesser propensity among the youth to use these types of participation. However, young people appear more likely to participate through cause-oriented activism than adults. In the case of multi-activism, it seems that age does not have a significant effect. These findings fit with the conclusions of most studies on participation and youth that indicate that young people tend to participate more than adults through protest and political consumption and less through more institutionalised forms. Regression model 2 in table 8 aims to see whether or not direct political involvement has a differential impact for young people with respect to adults. To find out, interaction between the direct involvement and youth variables is added. In this way, we can check if this new type of involvement has a more significant effect on the youth than on adults when explaining why they use more cause-oriented forms of participation. The results of this second model do indicate that the new political involvement variable has a special impact on the youth. To better interpret the differential effect that direct involvement has over young people and adults when explaining political activism, graph 1 shows the predicted probabilities of being active in different types of activism based on the new direct political involvement variable. The rest of the variables have remained constant 1. [Graph 1 about here] In model 1, the probabilities of participating politically in both cause-oriented activism and multi-activism increase when there is more direct involvement. However, the lines stay at the same level with institutional activism. This means that direct involvement has an affect on multi-activism and cause-oriented activism, but not on institutional activism. Regarding the differences between young people and adults, it is noted that being young increases the probabilities of participating in cause-oriented forms of activism and decreases with respect to forms of institutional-oriented activism. Regarding multi-activism, there are no significant differences between young people and adults. 1 The graph shows the probability depending on the variation of direct political involvement and the youth variable. The constant values that were given to the rest of the variables correspond to their central trend: average in the case of interest in politics (5.58) and mode in the cases of internal political efficacy (no) and external political efficacy (no), political discussion (yes), party identification (yes) and satisfaction with how democracy works (yes). 18

19 In the second model, when a variable of interaction is introduced that allows the increase to vary depending on young people and adults, there are changes with respect to the previous model. Logically, there are no differences in institutional activism given that direct involvement has no effect there. However, in cause-oriented activism and multi-activism, the positive effect of direct political involvement on participation disappears in adults and remains constant in young people. This indicates that the influence of this type of subjective approach to politics on activism is genuinely a feature of the young. These findings indicate that the effect of a subjective type of political involvement intervenes in young people s greater propensity to use forms of protest or political consumption. This involvement is not aimed at institutional stakeholders but directly at the social and political topics that arouse their interest. Moreover, this type of involvement is not covered by traditional survey indicators of political attitudes. 4. Conclusions In recent decades in Western countries, a series of transformations have occurred that have changed people s relationship with politics. This paper builds on the idea that surveybased studies into political behaviour have failed to detect one aspect of this change: the emergence of a subjective form of political involvement. In the absence of longitudinal data, it shows how today there is a type of involvement characterised by the fact that it is not oriented towards the traditional political mediating institutions or stakeholders of representative democracy. In addition to identifying this direct type of involvement and differentiating it from institutional involvement, the paper has demonstrated how the indicators of political attitudes commonly used in surveys do not capture this type of involvement, while doing just that for institutional-type of involvement. The identification of this different way of subjectively approaching politics helps to explain two phenomena related to these changes in people s relationship with politics. First, direct political involvement substantially improves the ability of political attitudes to explain cause-oriented activism. Behind the expansion and normalisation of protests (Norris, 2002; Ferrer and Fraile, 2007; Jimenez, 2011) and behind the emergence of political consumption as a form of political expression and participation (Michelletti et. al. 2003; Stolle, Hooghe and Micheletti 2005), there is a different way of feeling politically involved. Empirical models of 19

20 the influence of political attitudes on cause-oriented activism (which includes protest and consumption) improve substantially when the new variable of direct involvement is introduced. As such, it seems that most common indicators of attitudes of political involvement like interest in politics, internal and external political efficacy and party identification do not adequately cover a component of political involvement that is behind some of the most notable changes in people s political participation. Second, the existence of direct political involvement different from involvement aimed at institutional political objects helps to improve explanations and understanding of the youth s relationship with politics. Various qualitative studies have pointed to the fact that young people feel distant from party and institutional politics, but that does not mean that they don t follow politics in a broad sense (Henn et. al, 2002; O Toole et. al., 2004; Manning, 2010). Survey studies had shown that young people tend to use political protest and consumption more than adults (Norris, 2004; Zukin et. al., 2006). The findings presented indicate that this greater propensity of young people to use cause-oriented activism may be due at least in part to the fact that it entails a direct type of involvement that has an important effect on their forms of participation. The identification of direct political involvement helps us to achieve a more complex understanding of the subjective sphere of people s relationship with politics, and especially the youngest generations. The fact that classic survey indicators basically cover involvement of an institutional type may have contributed to overstating the phenomenon of the political disaffection of young people. Despite the fact that both types of involvement are not mutually exclusive, as we have seen, there are individuals that show strong levels of direct involvement while they don t feel interested in or close to the most institutional objects of politics. In these cases, we would need to speak of institutional disaffection rather than political disaffection. Moreover, these analisys have made it clear that direct involvement is closely linked with cause-oriented activism. Given the fact that such activism is increasing in most Western countries, one might think that direct political involvement is also increasingly important to the people of those democracies. 20

21 Table 1. Factor analysis on interest in political issues Components Immigration 0,667 0,221 0,126 Elections 0,155 0,793 0,122 North-South inequalities 0,700 0,180 0,200 Political parties 0,140 0,831-0,034 The rights of homosexuals 0,693 0,026 0,109 The relationship between Catalonia and Spain 0,088 0,488 0,382 The economy 0,165 0,324 0,640 Security -0,024 0,186 0,777 The environment 0,522 0,109 0,410 Social policies 0,591 0,452 0,050 The European Union 0,168 0,608 0,285 Housing 0,349-0,003 0,637 Access to employment 0,396 0,012 0,564 % variance explained (53,9%) 34,1% 11,0% 8,8% Rotation method: Varimax with Kaiser (6 iterations) Source: Enquesta de participació i política Generalitat de Catalunya. Table 2. Correlation between interest in politics and involvement indicators Interest in politics Direct implication Institutional Implication Basic implication Pearson s correlation 0,240 0,538-0,117 Sig. (bilateral) 0,000 0,000 0,000 N Source: Enquesta de participació i política Generalitat de Catalunya. Table 3. Implication indicators average score by political attitudes Direct implication Institutional Implication Basic implication N Internal political efficacy 0,100 0,060-0, External political efficacy -0,153 0,397-0, Political discussion 0,221 0,204-0, Party identification 0,000 0,307 0, Satisfaction with how democracy works 0,038 0,153-0, Source: Enquesta de participació i política Generalitat de Catalunya. 21

22 Table 4. Activism categories Institutional-oriented activism Causeoriented activism Activist Non activist Activist Multi-activist -10,9% ; N=227- Institutional-oriented activist (pure) -14,3% ; N=298- Source: Enquesta de participació i política Generalitat de Catalunya. Non activist Cause-oriented activist (pure) -13,7% ; N=286- Non activist -61,2% ; N= Table 5. Implication indicators average score by activism Direct implication Institutional Implication Basic implication Non activist -0,189-0,097 0,125 Institutional-oriented activist (pure) -0,013 0,275-0,036 Cause-oriented activist (pure) 0,350-0,038-0,215 Multi-activist 0,563 0,188-0,341 Source: Enquesta de participació i política Generalitat de Catalunya. 22

23 Table 6. Regression models on political activism by involvement attitudes Institutional-oriented activism (pure) Cause-oriented activism (pure) Multi-activism Model 1 Intersection -2,643 *** -2,72 *** -4,756 *** Interest in politics 0,105 *** 0,083 ** 0,289 *** Internal political efficacy 0,000 0,205 0,297 * External political efficacy 0,498 ** -0,484 * -0,039 Political discussion 0,784 *** 1,552 *** 1,676 *** Party identification 0,547 *** -0,142 0,355 ** Satisfaction with democracy -0,305 ** -0,277 * -0,498 ** N 2100 Pseudo-R 2 Nagelkerke 0,227 Intersection -2,34 *** -2,675 *** -4,636 *** Interest in politics 0,079 ** 0,075 ** 0,267 *** Internal political efficacy 0,001 0,153 0,148 External political efficacy 0,389 * -0,434 0,015 Political discussion 0,739 *** 1,472 *** 1,572 *** Party identification 0,539 *** 0,052 0,534 *** Satisfaction with democracy -0,402 *** -0,302 ** -0,563 *** Direct implication 0,111 0,438 *** 0,604 *** Institutional implication 0,093-0,160 * -0,111 Basic implication -0,151 * -0,307 *** -0,476 *** N 2100 Pseudo-R 2 Nagelkerke 0,276 Category of reference: Non activist Sig. ***<0,01, **<0,05, *<0,1 Source: Enquesta de participació i política Generalitat de Catalunya. Model 2 Table 7. Activism type by young and adults Young people (15-29) Adults (30 and more) Total Non activist 61,1% 61,2% 61,2% Institutional-oriented activist (pure) 8,5% 15,8% 14,3% Cause-oriented activist (pure) 19,6% 12,1% 13,7% Multi-activist 10,8% 10,9% 10,9% Source: Enquesta de participació i política Generalitat de Catalunya. 23

24 Table 8. Regression models on political activism by involvement attitudes and youth Model 1 Institutionaloriented activism (pure) Cause-oriented activism (pure) Multi-activism Intersection -2,873 *** -2,412 *** -4,233 *** Youth (living independent) -0,533 ** 0,487 ** -0,176 Interest in politics 0,105 *** 0,098 ** 0,235 *** Internal political efficacy 0,029 0,360 * 0,320 External political efficacy 0,468 * -0,192-0,283 Political discussion 0,666 *** 1,327 *** 1,340 *** Party identification 0,641 *** -0,036 0,519 ** Satisfaction with democracy -0,331 * -0,421 * -0,569 ** Direct involvement 0,069 0,454 *** 0,653 *** N Pseudo-R 2 Nagelkerke 0,275 Intersection -2,864 *** -2,547 *** -4,415 *** Youth (living independent) -0,518 ** 0,344-0,376 Interest in politics 0,104 *** 0,097 ** 0,234 *** Internal political efficacy 0,030 0,369 * 0,324 External political efficacy 0,466 * -0,211-0,293 Political discussion 0,669 *** 1,342 *** 1,341 *** Party identification 0,640 *** -0,034 0,522 ** Satisfaction with democracy -0,327 * -0,399 * -0,551 ** Direct involvement 0,024-0,296-0,036 Interaction youth*direct inv. 0,030 0,588 ** 0,575 * N Pseudo-R 2 Nagelkerke 0,281 Category of reference: Non activist Sig. ***<0,01, **<0,05, *<0,1 Source: Enquesta de participació i política Generalitat de Catalunya. Model 2 24

25 Graph 1. Predicted probability of different types of activism according to direct political involvement for young people and adults Source: Enquesta de participació i política Generalitat de Catalunya. 25