Since the earliest studies in electoral behaviour, research in sociology and political science has accumulated evidence that young people relate to politics differently than adults. Regarding their political participation, young people are less likely to cast a vote in elections (e.g. Blais, Gidengil and Nevitte 2004; Franklin 2004); they are less interested in politics in general (Martín 2004; García-Albacete 2014), and particularly detached from party politics and traditional institutions (Scarrow 2000). However, young citizens are also more willing to take the streets to make their voice heard (Barnes and Kaase 1979), and more likely to use creative forms of action (Norris 2002; Stolle, Hooghe and Micheletti 2005). When it comes to other political attitudes, they have been said to be more progressive and liberal (Glenn 1974), more ‘post-materialistic’ (Inglehart 1977; Inglehart and Welzel 2005), and more distrustful towards democratic institutions (Norris 2002; Dalton 2013); they are also simultaneously more supportive of norms of citizenship related to active engagement (Dalton 2008; Dalton 2013) and involved in their communities (Zukin, et al. 2006).

The arguments used to explain young people’s distinct political behaviour can be grouped in two main approaches. On the one hand, the life cycle explanation focuses on life stages and their relationship to politics. Each life stage that an individual goes through implies distinctive social roles, which confer different resources, interests, needs and demands to face the costs of being involved in politics (Jennings 1979; Sears and Levy 2003). Young individuals might be less aware of the importance of politics than adults, due to their lack of political experience. As they adopt adult roles (leaving the parental home, finding a first job and finishing their studies, marrying and having the first child), individuals become aware of the importance of politics in their daily lives and acquire the knowledge that makes involvement in politics easier and meaningful (Rosenstone and Hansen 1993; Strate et al. 1989). The relationship to politics becomes more distant again as people age, as retirement implies a decrease in social connections and exposure to mobilization networks and, finally, health issues (Milbrath 1965; Strate et al. 1989; Goerres 2009).1

On the other hand, age also has a psychological effect that is linked to the period in which citizens socialize; that is, the moment when individuals interiorize social norms and develop the basic attitudinal tools to interact with the political system (Hyman 1959; Jennings 2007). As young people set about developing their political views, they are more likely to be influenced by social transformations, and particularly so in critical periods (Ryder 1965). As individuals approach adulthood, their attitudes turn more resistant to change. The main theoretical consequence of the political socialization process is that the context in which individuals develop their political attitudes during their youth ends up conditioning their relationship to politics in the long term (Kinder and Sears 1985; Sears and Valentino 1997; Dinas 2013).

When it comes to explaining the impact of the economic crisis, both approaches can be useful. On the one hand, the crisis may have a significant impact on young people’s transition to adulthood, by halting, delaying or making more difficult the key transitional events: leaving the parental nest, finding a job, or forming a family (Sears and Levy 2003; García-Albacete 2014). On the other hand, the context of the crisis can be understood as a critical juncture in conditioning and influencing political views in the long run.

In light of previous studies on youth’s political behaviour, it is not surprising that young people were among the protagonists of street mobilizations fueled by an economic crisis that affected political systems all over the world. The full consequences of the crisis are still to be studied, but a degree of consensus has emerged amongst scholars. Regarding political attitudes, the crisis provoked a rapid decrease in levels of trust towards the main political institutions at national level (Torcal 2014; Roth, Nowak-Lehmann, and Otter 2013), but also at regional and European level (Armingeon and Ceka 2014). Although democratic legitimacy has not been threatened (Teixeira, Tsatsanis, and Belchior 2014), the sharp increase in dissatisfaction with how democracy works might be an expression of discontent about how democracies dealt with the crisis (Armingeon and Guthmann 2014).

Negative attitudes, however, did not lead to apathy or alienation. Protesters took to the street in 2011 to the point that they became the protagonists from Wall Street to Tarhir Square or Puerta del Sol, and across a number of other European Countries. The exceptional political mobilizations challenged previous theoretical accounts on who, when, how, and why people - and particularly the young - mobilize politically. Among the many theories developed to explain political behaviour in democratic societies, the Civic Voluntarism Model has proved to be the most effective in explaining who participates in politics. Participants are those who have high levels of resources such as time, education, money and/or civic abilities (Verba, Schlozmann and Brady 1995). Following this argument, recent studies provide convincing evidence that unemployment not only depresses political interest in early adulthood, but it also has scarring effects on levels of political engagement later in life (Emmenegger, Marx and Schraff 2017). Similarly, recent research on the political effects of labour market disadvantages shows that increases in both perceived job and financial insecurity erode trust in politics (Schraff 2018). Compelling evidence that greater economic inequality yields greater political inequality is also available at the aggregate level, as higher levels of income inequality powerfully depress political involvement among all but the most affluent (Solt 2008). Since the crisis reduces the availability of resources, we might expect a decrease in levels of political participation. However, if that is the case, how can the crisis-related protests be explained?

Given the limited capacity of resource-based theories to account for the intensity of protest movements during the Great Recession, some researchers moved to grievance theories to explain 2011 protest movements. While the Civic Voluntarism Model argues that material resources are an essential requirement for participation, the grievance model argues that relative deprivation will lead to more participation. As Kern et al. (2015) concluded, grievance theory is especially useful in exceptionally negative conditions since suddenly imposed grievances can lead to different modes of protest behaviour. In fact, the economic crisis diminished the traditional gap in terms of individual socioeconomic positions, at least in what concerns low-cost modes of political action (Rodon and Guinjoan 2018). Is also considerably boosted the public’s knowledge of the state of the economy and reduced information inequalities (Marinova and Anduiza 2018).

Finally, beyond individual approaches, it is relevant to take a look at the structure of opportunities found in each country. As Grasso and Giugni (2016) show, individual-level feelings or relative deprivation have a direct effect on the propensity to participate in protests, but that effect is greater in contexts characterized by either higher unemployment aggregate rates or higher levels of social spending; that is to say, in contexts that opened up political opportunities for protest among those who experienced a deterioration in living standards during the economic crisis. Bassoli and Monticelli (2018), when focusing on precarious young workers in five European cities, provide a similar argument. ‘Precarious youth’ are more active than regular workers –or not—depending on labor market regulations, thereby functioning as an issue-specific political opportunity structure. Interactions between individual level economic circumstances and the political context have also been identified in terms of government performance. How governments have handled the economic situation and the issue of unemployment alters the extent to which perceptions of egocentric economic situation affect political participation (Giugni and Grasso 2017).

The effect of other variations in the structure of opportunities has been less studied, for various reasons, among them the difficulties in identifying causal links. Institutional and contextual characteristics vary simultaneously across countries and over time. In the early years of the crisis, almost all developed countries suffered its consequences in terms of slower economic growth, higher unemployment, and problems balancing the public deficit (Matsaganis and Leventi 2014). However, since 2011, differences among developed countries can be observed within Europe. As stated in the introduction, Northern European countries such as Denmark, The Netherlands and Germany overcame the worst consequences of the crisis around 2011. On the other hand, Spain, Portugal, Italy and Greece’s economic situation worsened in 2012, when levels of unemployment reached rates never seen before (Braun and Tausendpfund 2014). Political outcomes also varied between European regions. Even though protests spread across almost all developed countries (Bermeo and Bartels 2014), electoral shifts were very different between Northern and Southern European countries (Hernandez and Kriesi 2016; Bosco and Verney 2012, 2016). In Southern Europe party systems underwent major transformations due to crises among the mainstream parties and the emergence of new political formations in Greece (2012), Italy (2013) and Spain (2014/2015) (Teperoglou, Tsatsanis, and Nicolacopoulos 2015; Orriols and Cordero 2016; Morlino and Raniolo 2017). The only exception was Portugal, where the party system remained stable (Fernandes, Magalhães, and Santana-Pereira 2018).

Overall, the structure of opportunities interacted with social demands and shaped citizens’ behaviour. In both protest movements and in the electoral success of new political parties, young people played an important role (Della Porta et al. 2017). Thus, young people’s relationship to politics changed significantly during those years. But, what has happened since then? To the best of our knowledge no study has provided a comprehensive view of levels of political engagement and mobilization in the aftermath of the Great Recession. Following our previous discussion, we offer three different proposals to respond to the research question.

Firstly (1), given the politicized context of the crisis in which this youth cohort came into politics, we expect that they may exhibit higher levels of political involvement after the crisis. This expectation is supported both by both socialization and grievances theories. Secondly (2), the literature suggests that young people are no longer as involved in politics as they once were. A decade after the economic crisis, the economic circumstances of young people are still challenging. They still suffer precarious conditions that hamper their ability to face the costs of being politically engaged. Consequently, the alternative expectation is that young people may exhibit lower levels of political involvement after the crisis. Finally (3), following country-specific characteristics, political involvement and participation can be sensitive to the particular structure of political opportunities. Therefore, we expect that young people from countries particularly hit by the crisis will remain more politically involved than their counterparts in those that were less affected by the Great Recession.

As said, in this section we present our descriptive analyses of young people’s political attitudes and behaviour over time. Findings show great heterogeneity across countries. In order to make interpretations easier, in what follows, for each attitude or behaviour we first address the general trend for the entire population and then we establish a double comparison: (a) between the youth and adults over time, and (b) among young people before, during and after the economic crisis.

When looking at the trends for the whole population, there is a large variation in levels of political interest over time in each country (see Figure A1 in the Appendix). In some countries, the economic crisis came with an increase in political interest. That is the case of Germany, Spain, France, Great Britain, Ireland and Slovenia when we compare average citizens’ political interest in 2006 to 2008 or 2010. In other countries, the crisis did not have a visible effect, and levels of political interest remained stable (Belgium, Switzerland, and Norway). In some countries the increase in political interest in 2008 was followed by a visible decrease in political interest shortly after (2010-2012), such as in Ireland and Great Britain. A decrease in political interest is also visible in Hungary around the same time. In two Nordic countries, Sweden and Finland, the increase in political interest is only visible later on, in 2014-2016. In those cases, the increase takes place far beyond the hardest years of the economic crisis and might be related to another crisis, in this case, the so-called refugee crisis.

Regarding age differences, Figure 1 shows the average level of political interest for each age group (young people in black, adults in grey and seniors in white) by country and over time (2002-2016). A common pattern emerges: young people are significantly less interested in politics than adults (as represented in black lines). This is not surprising, as previous research has repeatedly shown that young people are less interested in politics (e.g. Glenn and Grimes 1968; Prior 2010). This gap can be understood with the life-cycle hypothesis discussed above, and particularly, by a lack of experience in – and knowledge of – politics. Only Portugal, where young people are as interested in politics as adults, deviates from this pattern. When compared to other South European countries such as Spain, the cohort of young people socialized in democracy has maintained relatively high levels of political interest, due to the character of the transition to democracy (Fishman and Lizardo 2013). An additional explanation could be that Portugal suffered an earlier economic crisis in 2002 that mobilized young people significantly.5

Regarding the evolution of young people’s political interest over time, we can observe some differences in comparison to adults, but also variation across countries. Young people and adults follow the same pattern in Germany, France, Great Britain and Ireland, where we observe an increase in their levels of political interest in 2008, when the crisis began. In Switzerland and Norway, there are no changes over time. In four countries, Belgium, Czech Republic, The Netherlands and Slovenia, increasing or stable levels of political interest among adults contrast with negative trends for young people. Overall, and in line with the ‘impressionable years’ thesis that emphasizes how young people are more sensitive to contextual changes, we find larger variations across time for young people than for adults. Adults’ levels of political interest remain more stable over the years, without sharp changes, while young people’s levels of interest show abrupt changes in short periods. Figure 1 exemplifies such changes in Finland, Great Britain, Ireland, The Netherlands, Norway, Sweden and Slovenia, where significant ups and downs are visible for young people even in short periods of two years.

Let’s now turn our attention to confidence in institutions and specifically to levels of trust in the country’s parliament (see Figure A2 in the Appendix). Once more, we observe large variation across countries. A decrease in trust during the crisis is found in eight countries, and no changes in five cases; while in only two countries the level of trust increased during the crisis. The countries where trust has decreased since the crisis (2008-2010) are: Germany, Spain, Finland, France, Great Britain, Ireland, Portugal and Slovenia. We find no changes in trust in Belgium, Switzerland, Czech Republic, The Netherlands and Norway. Finally, we observe an increase in the levels of trust in the national parliament in Hungary, Poland and Sweden.

Young people do not distrust political institutions any more than adults or senior citizens do (see Figure 2), contrary to what other studies have claimed (Norris 1999; O’Toole, Marsh and Su 2003; Dalton 2004: 94; Sloam 2007)6. This finding is not exceptional, as previous studies have provided evidence that young people trusted basic institutions of democracy as much (if not more) than adults at the beginning of the 21^st century (García-Albacete 2014: 142-143). As the black lines indicate in Figure 2, on average, young people declare higher levels of trust in national parliament in almost all countries at some point during the period 2002-2016 (Belgium, Switzerland, Czech Republic, Germany, Finland, France, The Netherlands, Portugal and Sweden). In other countries (such as Spain, Great Britain, Hungary, Ireland, Norway, Poland and Slovenia) we find no differences between adults and young people. Over time, young people and adults follow the same trend of stability. This is clearly visible in Figure 2, where lines overlap with each other.

In sum, the descriptive evidence does not allow us to determine whether young people are particularly interested in politics or critical towards institutions. Both adults and young people share the same trends. We expect to observe more differences regarding political behaviour.

Let us now turn our attention to different modes of political participation. Figure 3 shows the percentage of young people that participate only in protests (black), only in institutional politics by voting (white), in both protest and institutional actions (light grey); and finally, the percentage of young people who do not participate in any activity (dark grey). Figure 4 provides the same information for adults.

The first visible pattern is common to both young people and adults. Among both age groups there is a larger percentage of citizens who participate exclusively in electoral politics. Notwithstanding this regularity, there are also visible differences between age groups. Young people participate to a greater extent than adults in protest or in both institutional and non-institutional actions. Regarding non-participants, the percentage of young citizens who do not participate in any kind of political action is also larger than the percentage of adults who do not participate. This evidence confirms something that other studies previously found: young people are less participative in electoral politics (Blais, Gidengil and Nevitte 2004; Franklin 2004) and when they participate, they prefer non-institutional forms of action (Barnes and Kaase 1979; García-Albacete 2014). Finally, the relatively large percentage of young citizens in the mixed category ‘both’ confirms that they are more likely to combine different forms of political participation (García-Albacete 2014).

Having identified general patterns and differences between young people and adults, we now focus on trends over time. First, Figures 3 and 4 show that although there is large heterogeneity across countries, adults and young people’s participation followed the same dynamics. That is to say, when percentages of any mode of political participation increase in one specific year, this increase is typically found for both young people and adults. The same occurs when political participation decreases. Same trends across age signal the existence of period effects that have had an impact on all citizens.

Even though the period effects are shared by young and adult citizens, we now focus on changes across time and particularly after the crisis started (see in Figures 3 and 4, from 2008 to 2012). In some countries, institutional participation remained stable, for both adults and young people (see for instance Belgium, The Netherlands, Ireland and Norway). In a second group of countries, institutional participation decreased; although the change is less intense among adults, as in the case of the Czech Republic and Finland. In a third and more numerous group of countries, institutional participation increased; this is the case of Germany, Spain, Hungary, Poland, Portugal, and Slovenia. Finally, in countries like Switzerland, France and Sweden the trends shift during the period of crisis.

Regarding other types of participants, there are no clear trends during the crisis. In some countries - Belgium, Finland, France, Great Britain, Hungary, Ireland, The Netherlands, Sweden and Slovenia - the percentage of respondents that both participated in institutional and non-institutional actions, and the percentage of respondents that only protested, decreased. In other countries, such as Switzerland, Germany, Spain, Poland and Portugal, we can identify an increase in the levels of political participation. Additionally, in these countries, the increase in protest occurs instead in the post-crisis period.

The increase in political participation from 2012 onwards is a common trend for both adults and young people (see Figures 3 and 4). It is visible for all types of participation and it has its parallel in a decrease in the percentage of respondents that did not participate at all. It is, however, in the post-crisis period (2014-2016) when we find more political involvement: the number of citizens that participate in institutional actions and in both protest and institutional modes increased significantly. To provide a better illustration, Figure 5 focuses on the percentage of non-participants (category “none”) for young people (black lines) and for adults (grey lines).

Again we observe a general trend despite heterogeneity across countries: in the post-crisis scenario there is greater mobilization. At the end of the period only in the Czech Republic, France and Hungary is there a larger percentage of respondents who are not involved in politics than during the crisis. As we previously found, with regards to other indicators, young people are more volatile, and thus tend to change more in a shorter period of time. In addition to the general trend of political mobilization, Figure 5 shows that growing mobilization particularly affected young people. Adults did not change in many countries or, as happened in Hungary and Slovenia, followed a different pattern than young people.

In sum, young people, who are more sensitive to context, changed their political behaviour to a greater extent than adults. This change is consistent with our first expectation: that young people became more politically involved during the crisis. Despite experiencing hard economic conditions, the fact that they socialize politically during a politicized period resulted in increased levels of political participation. This mobilization finds no parallel among adults, signalling that the cohort effect might be a good mechanism to explain the changes observed. Regarding our third expectation that changes would be particularly visible in the countries hardest hit by the economic crisis, the descriptive findings do not support this view. Further analyses are required to understand these differences better. We address this issue in the next section.

The descriptive findings support the expectation that citizens became more politically involved during the crisis and maintained high levels of political involvement thereafter. To confirm the robustness of the results, controlling for the potential different composition of age groups, we run logit equations conducted separately for each country. Figure 6 summarizes the results by plotting the predicted probabilities of not participating in politics (versus any form of political participation) over years and across countries. We have introduced as the main independent variable the interaction of age and year and, as control variables, gender, level of education and occupational status (the data are available in Table A1 in the Appendix). The results confirm previous descriptive findings. Whereas adults’ participation has remained fairly stable, young people’s participation increased over time. We observe that young people’s levels of non-participation (higher values) decrease over time in Belgium, Switzerland, Spain, Finland, Norway, Poland, Portugal and Slovenia. On the other hand, we see stability for both adults and young people in Germany, Great Britain, Ireland, The Netherlands and Sweden. Finally, young people tend to be less participative in recent years in the Czech Republic, France and Hungary.

Again, we interpret higher levels of involvement as a consequence of a highly politicized period after the economic crisis. Far from creating distance from politics, we see that one long-term effect of the crisis is an increase in young people’s political engagement. Although political attitudes changed in a similar way for adults and young people, among young citizens interest translated into political action to a greater extent than among adults. It could be that the high levels of political interest and the acquisition of participatory habits become a cohort characteristic that could have long term effects.

Following previous arguments, we also expected this pattern to differ according to the country-specific structure of opportunities. Specifically, we expected mobilization to be greater in countries that were particularly affected by the crisis. However, as we discussed before, no trend can be found across countries that supports such expectation. Over time, levels of political participation have increased in countries that were harder hit by the crisis, such as Spain and Portugal; but also in countries where the crisis was less intense, such as, for instance, Germany, The Netherlands, Finland and Norway.

Our research design is not well equipped to explain this finding as there are several factors that may have influenced young people’s high levels of political involvement in recent years. High participation levels could be related to country specific domestic politics, such as whether elections were particularly competitive (as competitiveness in elections is a good mobilizer for young people; see Franklin 2004) or the success of mobilization tactics by new or small political parties. An increase in political participation might also be the result of international political conflicts beyond the Great Recession, such as the refugee and migration crisis. Most likely, both explanations account for the increase in young people’s political participation. The economic crisis might have played a greater role in Portugal and Spain, whereas the reactions to migration politics might explain mobilization in Central and Northern Europe. Economic and migration crisis might also interact.

Irrespective of the specific events that have increased political participation, the increase in mobilized citizens over time and across countries allows us to be optimistic. On average, young people are, after the crisis, more participative than before, in spite of the strategies followed by certain governments to increase the costs of participating in politics.