Emotions and climate strike participation among young and old demonstrators

ABSTRACT Research shows that anger triggers participation in social movements, while fear inhibits action. Therefore, fear is less likely to contribute to citizens’ engagement in protest. However, in the case of climate change, fear may play a distinct role and thus contribute to participation. Given the long-term consequences of climate change, we argue that it triggers different emotions across disparate age groups. We investigate the extent to which young, adult, and senior climate strikers experience fear and anger in relation to climate change. Furthermore, we analyze the contribution of these emotions to younger and older citizens’ motivation to demonstrate. Using a unique dataset collected among climate strike demonstrators in eleven cities around the world in September 2019, we examine the importance of anger and fear in explaining motivations to take part in the demonstration – to pressure politicians or to defend one’s interests. Overall, we find that protesters aged above 60 years old are less likely to fear climate change but are more likely to feel anger in relation to this issue than younger generations. On the other hand, those aged below 35 report the highest levels of fear in relation to climate change and are significantly less angry than senior citizens. In all age groups, both anger and fear are associated with motivation to defend one’s interest and to pressure politicians.


Introduction
The climate strike movement led to thousands of people taking to the streets to demand action against climate change.Young people initiated this movement and, as they repeatedly demonstrated during 2019, older citizens joined the movement (Lorenzini et al., 2021).Research on these strikes has investigated the composition of the strikes, the motivations of strikers, and their tactics and targets (de Moor, De Vydt, et al., 2020;de Moor, Wahlström, et al., 2020;Wahlström et al., 2019).Furthermore, research on social movements has shown the importance of emotions in political participation (Goodwin et al., 2001;Hochschild, 1979;Jasper, 2011;Whittier, 2021).More specifically in relation to climate change, research has argued that this issue triggers a variety of emotions (Kleres & Wettergren, 2017;Poma & Gravante, 2021).It has also shown that threat related emotions, such as fear (sometimes referred to as eco-anxiety), and other negative emotions, such as anger (sometimes referred to as eco-anger), are the two emotions that are most often referred to by climate strike protesters and are potentially relevant for their participation (Pickard, 2021).
Previous studies on emotions and social movements have shown that anger fuels participation (Jasper, 2014(Jasper, , 2018) ) whereas fear tends to inhibit it (Azab & Santoro, 2017).However, the impact of fear depends on the context and the issue which requires mobilization.Fear of repression leads to inaction (Young, 2016), whereas fear of stigmatization requires emotional management to successfully contribute to protest (Hochschild, 1979).Fear of losing access to vital resources contributes to mobilization (Gravante & Poma, 2016;Simmons, 2014).Regarding climate change, fear relates to worries about the future and the conditions under which oneself, future generations, or others around the world will live in years to come.This kind of fear resembles that of losing resources, which triggers action.In the context of the Great Recession, studies have found that people who fear a worsening in living conditions experience grievances that trigger action (Galais & Lorenzini, 2017;Portos, 2020).
In this paper, we argue that, in relation to climate change, both fear and anger foster motivation to protest.We examine whether young and old climate strikers feel different emotions towards climate change and whether these emotions have disparate impacts on their motivations to protest.We therefore seek to identify the effects of emotions on motivations to protest among different age categories of climate strikers.We focus on two important motivations: to defend one's interests and to pressure politicians.Therefore, we contribute to a more nuanced understanding of climate strike activism and highlight its heterogeneity.
Research has found that environmental issues impact mental health (Bourque & Cunsolo Willox, 2014) and well-being (Hrabok et al., 2020).Climate change poses many challenges to the well-being of both current and future generations.Therefore, it raises important questions associated with inter-generational solidarity and equity.For instance, it aids in the examination of to what extent current generations need to reduce their level of well-being today to ensure some level of well-being in the future (Sachs, 2015).In addition, climate change disparately impacts the well-being of young and old citizens.In Sweden, young people are more concerned about climate change (Uba, 2021).More generally, research shows that environmental concern triggers eco-anxiety especially among younger generations (Hickman et al., 2021).Children and young people are not only worried about climate change, but they are also heavily affected by adults' inaction (Hickman, 2020).They feel that there is a disconnection between the urgency of the situation and adults' prevailing inaction.Young people are more likely to feel worried.However, the youth have limited economic and political power to trigger social change.
Overall, research has found that eco-anxiety, fear, and anger all contribute to political action (Stanley et al., 2021).However, it is unclear whether social groups differ in their emotional responses to social issues and if, within a movement, different groups experience different feelings towards the cause.Therefore, we investigate if young, adult, and senior climate strikers equally experience anger and fear when thinking about climate change.We further consider whether these emotions similarly influence youth, adult, and senior citizens' motivations to participate in the climate strike movement.
During each month of 2019, in many countries around the world, young people organized climate strikes and marches (de Moor, De Vydt, et al., 2020).Some of these events were coordinated across countries.This was the case for the September 2019 week of action that we analyze in this paper.During this protest week, research teams in fifteen countries gathered data about demonstrators as they marched in the streets.We use the climate strike data from 11 demonstrations which took place across eight European countries, Australia, and the United States (see Appendix I for a list of countries and cities).This unique dataset allows us to compare young and senior demonstrators' emotional reactions to climate change and their motivations to take to the streets.
Our paper seeks to shed light on inter-generational differences in climate strike activism.As we work only on demonstrators, we cannot seek to understand participation.However, we can identify two different motivations to demonstrate.We compare young, adult, and senior climate strikers' motivations to defend one's interest and to put pressure on politicians.
In the remainder of the paper, we first present the role of emotions for protest participation and discuss how emotions are expected to differ across age groups.We then introduce the dataset collected across 11 demonstrations on climate change and the operationalization of the variables that we use in our models.Lastly, we present our empirical analyses.We conclude with a summary of our main findings.

Emotions and protest
Emotions play an important role in predicting protests and other social movements (Van Troost et al., 2013).In the last few decades, the study of emotions in social movement research has gained pace, building on the study of culture and focusing on the interplay between emotions, framing, and social identities (Jasper, 2011).Regarding framing, social movement organizations use emotions such as anger and pride to mobilize citizens around specific issues, collective interests, and identities.Furthermore, emotions play an important role constructing solidarity within a group.Reciprocal emotions within a group, such as joy and love, bind members of the group together.Furthermore, shared emotions towards outsiders reinforce these reciprocal emotions (Jasper, 2014).For instance, shared anger or outrage regarding the misbehavior of political authorities reinforces reciprocal emotions within the group.This may happen in relation to a moral shock.For example, this could occur when people discover that a reality is not as they believed it to be, producing a visceral unease and potentially driving action (Jasper & Poulsen, 1995).In this case, anger and indignation lead to political action.Similarly, positive emotions lead to a willingness to act and to engage in a political group, but also to take part in demonstrations that appear to be fun activities.Different emotions contribute to protest and often people experience multiple emotions simultaneously (Van Troost et al., 2013).
In this paper, we focus on the role of anger and fear in explaining participation in climate strike demonstrations.More specifically, we are interested in the issue-related emotions triggered in relation to a given protest issue.In our case, the issue is climate change, which may trigger anger regarding governments' inaction, or fear of the consequences of climate change for an individual's and others' living conditions.For older climate strikers, fear may not be associated with their personal experience of these living conditions but instead may be on behalf of their children or grandchildren and, more generally, for future generations.Below, we explain how these two emotions relate to protest.

From anger to indignation and protest
James Jasper (2014, p. 208) writes that 'anger and indignation, the morally grounded form of anger, are crucial to many aspects of protest'.Anger is considered an engaging emotion, and one that will lead to action (Berkowitz & Harmon-Jones, 2004).Appraisal theory states that appraisal, emotions, and actions are available means to cope with events in individuals' social world (Scherer et al., 2001).Social movements build on anger to highlight mobilizing elements which are associated with injustice or blame, and which are likely to trigger action (Van Troost et al., 2013).Anger will result in action when people feel a sense of efficacy.They feel this when they believe that their political action will change the course of an event (Van Zomeren et al., 2004).In addition, collective identity plays an important role in relation to anger.Collective identity may interact with emotions to trigger action (Van Stekelenburg & Klandermans, 2013) or constitute a different path towards political action (Van Zomeren et al., 2004).The former considers that 'people who perceive the in-group as strong are more likely to experience anger and desire to take action [. ..]' (Van Stekelenburg & Klandermans, 2013, p. 893).In the latter, the authors argue that group-based anger and group efficacy are both supportive of collective action.Group-based anger refers to a focus on emotion, whereas group efficacy focuses on coping with a problem.
Many scholars believe that anger plays an important role in protest.'Anger moves people to adopt a more challenging relationship with authorities than subordinate emotions such as shame and despair or fear' (Van Stekelenburg & Klandermans, 2013, p. 893).Building on this, we expect that anger will play a more important role in predicting the motivation to set pressure on politicians than on the motivation to defend one's interest.

Collective action to cope with fear and seek agency
Fear can save an individual's life when it triggers reflex reactions such as fleeing from danger.In the context of social movements, fear is understood as something that can either paralyze or lead to outrage (Jasper, 1998).When confronted with fear, individuals can opt for one of two coping strategies.Firstly, the emotion-focused coping strategy seeks to get rid of the negative feeling (Ojala, 2016).This results in avoidance, distancing, and disengagement.For example, a Norwegian rural community collectively seeks to avoid the issue of climate change in their daily lives, despite irrefutable evidence that climate change is taking place and threatening their living environment.This is done to keep at bay fear, guilt, and helplessness (Norgaard, 2011).The second strategy is the problemfocused coping strategy.It highlights that people seek to confront a problem and to find solutions (Ojala, 2016).This involves seeking information, adapting one's lifestyle (e.g., cycling, consuming less, becoming vegetarian), and trying to influence others, such as family members and friends.
In the case of climate change, fear is considered conducive to action (Kleres & Wettergren, 2017;Stanley et al., 2021).Fear is associated with the consequences of climate change and inaction.Therefore, fear is external to the movement.These findings complement social movement research which focuses on fear within movement settings (such as fear of police repression) that leads to inaction and avoidance.Hochschild (1979) investigated the process of fear management.She defined this process as the transformation, reduction, or suppression of fear to mobilize group members who are afraid to participate due to fear of repression or discrimination.However, as is the case for climate change, fear may be generated outside of the movement (Azab & Santoro, 2017).In this case, there is no need to manage or reduce fear.Fear can even be a driver of action.For example, people who fear they will lose access to water develop grievances that are conducive to political resistance (Simmons, 2014).Gravante and Poma (2016) distinguish two types of fear: 'reflex' fear, which inhibits action, and 'moral' fear, which leads to action.Fear generated outside of a movement is more likely to be a moral fear that is conducive to action.
In the case of threats to people's living environment, such as those associated with climate change, fear relates to an imagined future that people do not wish to experience.In these cases, citizens may want to engage in protest to express their fear and draw attention to the issue that caused it.'People can protest because they are tired of living lives of fear.Fear undermines a person's sense of control over his or her life and protest participation is a way to reclaim agency' (Azab & Santoro, 2017, p. 478).This corresponds to the problem-focused coping strategy discussed above.

Age, emotions and motivations to attend a climate strike
The above discussion clarified the important role which emotions play in social movements.While the literature shows that emotions vary by issue and their effects are context dependent, it has not yet investigated the possibility that social groups feel different emotions in relation to an issue.In addition, scholarship has not considered the possibility that emotions play a disparate role in motivating participation depending on the characteristics of individuals feeling them.Emotions are likely to differ depending on personal relation to an issue.Regarding climate change, age is an important individual characteristic which must be considered.Indeed, age groups systematically differ on several attributes that are relevant for the emotions felt in relation to climate change.Moreover, the youth, adults, and seniors differ in their motivations in joining a political protest.
One of the differences between age groups is their life expectancy.Although adverse consequences of climate change are visible today, these are expected to worsen over the next few decades (IPCC, 2021).The public is aware of this and the elderly generations' lower likelihood of being personally affected by climate change is one of the main reasons that has been put forward to explain age differences in climate worries (Andor et al., 2018).Following this logic, we can expect young respondents to be particularly worried about climate change and therefore more likely to feel fear.
A second difference between age groups is their relationship with politics.Here, there are both life cycle and generational effects at play.The life cycle affects individuals' likelihood to participate in politics.Political participation in institutional political actions, such as voting, is particularly high among individuals in their 50s.Conversely, voting is lowest among the youngest voters and among the elderly (Bhatti et al., 2012).Participation rates in institutional politics reflect generational differences in political trust to some extent.Trust in government has been declining over the recent decades and, when compared to earlier generations, individuals born in the 1980s and later display significantly lower levels of political trust (Dalton, 2005;Twenge et al., 2012).
For these reasons, which can be summarized as self-interest and relation to politics, we expect that age groups experience different emotions about climate change and may have different motivations in attending a climate strike.
Firstly, regarding emotions, younger individuals can be expected to fear climate change more than senior individuals.The main reason for this expectation is that young individuals are more likely to be personally adversely affected by climate change during their life, due to their longer life expectancy.Previous research conducted on the American population has shown that younger individuals reported higher levels of climate change fear (Firebaugh et al., 2021).On the other hand, there are reasons to believe that anger is more prevalent among older climate strikers than younger ones.This is connected to older generations' higher expectations towards institutional politics.In addition, older generations may, as a result, feel particularly angry with slow institutional responses to climate change.In the case of younger generations, they are less likely to feel betrayed by a political class that they do not trust anyway.Existing research has shown that political trust has been declining during recent decades and is currently lower among younger than older cohorts (Dalton, 2005) and that younger individuals are more likely to feel anti-establishment sentiments (Droste, 2021).Furthermore, older strikers may have already taken part in previous waves of environmental protest and, therefore, may feel increased frustration that things have not progressed.This is likely to trigger anger about lost opportunities to prevent climate change.Therefore, the expectation is not that young activists are not angry.As they are likely to be affected by climate change, they are very likely to be angry.However, older climate strikers are likely to be slightly more angry because they grew up in an environment in which expectations towards politicians were high and are now confronted with the inaction of political representatives, which triggers other condemning emotions, such as anger.
Age groups are also likely to differ in the motivations that bring them to climate strikes.There are many different motivations for individuals to participate in climate strikes.Some may participate to feel a sense of community, whereas others may want to raise awareness about the issue or inspire other citizens.However, there are two explanations which are most commonly given to explain climate strike participation: to defend one's own interest and to pressure democratic representatives to take political action.
These two motivations highlight the goals of activists.Putting pressure on representatives implies having specific requests about collective action to be taken at the political level.Conversely, defending one's interest is a more general motivation that does not imply relying on state structures to resolve climate change.There are reasons to expect that the importance given to each of these goals will differ according to age.Due to their likelihood of being directly affected by climate change, we would expect young strikers to be more likely to be participating in the climate strike to defend their own interests.Furthermore, the older generation, who are less likely to suffer personally from climate change, will probably not consider defending their own interests as a major motivation.
Due to their generally higher expectations of political institutions (Dalton, 2005;Twenge et al., 2012), older strikers will be more likely to participate with the goal of influencing political elites.Arguably, someone who does not trust political institutions has little incentive to pressure political representatives as they may not be perceived as reliable anyway.Therefore, we expect that younger climate strikers are less likely to consider pressurizing elected politicians as a main motivation to participate in the climate strike.
Finally, the literature on emotions and participation in protests shows that emotions and motivations are likely related.Anger has been found to boost political participation.We can therefore expect it to be positively associated with factors which motivate protest.The angrier individuals feel, the more they will want to defend their self-interest and to pressure politicians to change the situation.Although fear inhibits political participation, it has been found to have the opposite effect in relation to environmental political actions.Therefore, in the case of a climate strike, fear is positively associated with motivation to take part in a protest.While we expect different levels of emotions among age groups, we do not assume that there should be differences in the extent to which these emotions affect motivations across age groups.This possibility will be tested through our empirical analysis.

Data and method
For this study, we use data collected at climate strike demonstrations from a large coalition of research teams across different countries (de Moor, De Vydt, et al., 2020;Wahlström et al., 2019).The data were collected during the September week of climate action in 2019.During each of these demonstrations, teams of students and researchers distributed up to 1,000 flyers to participants.The number of flyers distributed varied from 211 in Gothenburg (Sweden) to 1,000 in Bern (Switzerland), Florence (Italy), and Helsinki (Finland).For more information about the number of flyers distributed, see Appendix I.The flyer gave participants access to an online survey about their current participation in the demonstration, their political attitudes, their political behaviors, and socio-demographic control variables.The participants were selected using a procedure to randomize the selection and to construct a representative sample of participants in the demonstration (Walgrave & Verhulst, 2011).
It is important to note that the response rates varied between 12 and 34%.We observed the lowest response rate in Florence (Italy) and New York (USA), with respectively 12 and 13%.The response rate reached a third of the distributed flyers in Finland (34%).In five cities, about a quarter of the people who received a flyer took part in the survey.The response rate was 29% in Bern (Switzerland) and Sydney (Australia), 28% in Gothenburg (Sweden), and 27% in Berlin (Germany) and Vienna (Austria).The variation in response rates does not preclude the fact that more committed demonstrators answered the survey across the cities.
Indeed, climate strikers who took part in the survey were more interested in politics than average climate strike participants.We can identify such biases as when we handed out flyers to access the online survey, we asked one in five potential survey respondents to fill in a short online questionnaire to assess their sociodemographic profile, political attitudes, and the relation to protest of the people who had participated in the different demonstrations.This allows us to analyze the composition of the event and to gain knowledge about the population from which the sample was drawn.We can compare the online survey respondents to demonstrators in terms of their sociodemographic profiles (sex, age, and level of education), political attitudes (political interest), and relation to protest (prior participation in demonstrations, timing of the decision to participate in this demonstration).In so doing, we can assess the representativeness of the sample.Comparing survey respondents to street demonstrators, we identified one systematic bias: those demonstrators who are interested in politics are more likely to answer the online survey than participants invited to fill in the web survey who are not interested in politics.This confirms that, despite the variation observed in response rates, the identified bias was the same across the different countries.
This data collection method offers a unique opportunity to analyze demonstrators who participated in specific mobilizations.Therefore, the data enables us to analyze the characteristics of demonstrators based on their effective participation (as opposed to reported participation) and to understand the specific issue of the mobilization in which they participated.However, this comes with significant disadvantages.This data collection method required a large team of trained data collectors to hand out flyers during the demonstrations and to obtain a representative sample of all participants.In addition, like other survey methods, there are biases associated with the profile of those who will answer, or will not answer, a research survey.Although we have controlled for sociodemographic biases and political interest, other biases are also important.For instance, individuals with specific characteristics, such as having young children at home, may have had less time available to answer the survey.Similarly, some senior citizens may not have felt comfortable using a computer, tablet, or phone to fill out the questionnaire.However, we anticipate that this bias will not have been as strong because demonstrators who take part in this kind of environmental movement are more likely to have high levels of education than other citizens in their age category.

Group categories
We define senior citizens as all the climate demonstrators who are 60 years old or over.We set the threshold at 60 years old for two reasons.First, research on political participation has found that participation in political activities other than voting starts to decline after 50 (Melo & Stockemer, 2014;Quintelier, 2007).However, we set the threshold closer to the age of retirement.This is our second reason, as in many countries retirement age is increasing (average age is 63.5 for women and 64.2 for men in OECD countries 1 ).However, in some countries, like Sweden and Australia, it is still possible to take an early retirement between 55 and 60 years old.Therefore, we decided to set the threshold at 60 to include a large share of pensioners and to capture the age where political participation is already declining.
We compare senior citizens to young climate strikers who are aged 15 to 34 years old to adult strikers who are aged 35 to 59 years old.We set the relevant age categories in accordance with a lifecycle perspective ( Van de Velde, 2008).Within this understanding, adults are individuals who are financially independent, live on their own, or with their own nuclear family (i.e., partner and/or children).As new generations may engage in longer education or face more difficulties in entering the labor market and gaining a stable income, following other research on youth political participation, we set the threshold at 34 years old (Giugni & Lorenzini, 2017;Giugni et al., 2021).Appendix II presents the distribution of respondents by age categories in the ten countries.

Countries included in the analyses
Since we seek to compare young, adult, and senior climate strikers, we selected countries that include a sufficient share of senior citizens.We investigated countries with more than 10% of senior citizens to have enough respondents that fall within the older age category in each country.This means that we analyzed data on 11 demonstrations collected in 10 countries.In the regression models, we controlled for countries using fixed effects.

Dependent variables
We analyzed two dependent variables that capture motivations for taking part in climate strike demonstrations.We asked the survey respondents to 'please indicate to what extent you agree or disagree with the following statements: I participated in the demonstration in order to. ..' and we use the following two items: 'defend my interests' and 'pressure politicians to make things change'. 2 For each item, respondents could answer on a fivepoint scale ranging from 'strongly disagree' to 'strongly agree'.We recoded the responses to construct dichotomous variables, distinguishing between those who strongly agree with the motivation to participate from all others.The choice of recoding these variables was linked with the fact that the distribution was skewed, with a very small share of respondents choosing the first two response categories (between 2 and 8%).We ran a correlation between the two dependent variables to ensure that they are independent and measure different concepts.The correlation between defending interests and pressuring politicians was low (r = 0.27).

Independent variables
Our main independent variables measure emotions, building on a question that asked 'thinking about climate change/global warming makes me feel' and gave the options 'fearful' and 'angry'.In each case, the respondent could answer on a five-point scale ranging from 'not at all' to 'very much'.Descriptive analyses showed an imbalance in the number of respondents in each of the five categories, with many more respondents in the last two categories than the first three.Therefore, we decided to recode the variables into three-state categorical variables to maintain these nuances, instead of dichotomizing the variables.We therefore created two three-state categorical variables, coded as 0 = not fearful/angry; 1 = quite fearful/angry; 2 = very much fearful/angry.

Control variables
We controlled for sex of the respondents, subjective social class, 3 and political interest.These variables have been found to systematically affect political participation and are likely to differ across age categories.
We measured sex of the respondents with a dichotomous variable, distinguishing between male and other respondents.The original variable included 51.9% women and 1.8% of respondents who selected the other category.We grouped women and those who selected the other category, including non-binary, gender fluid, and other subordinate positions in patriarchal society.
We use a subjective measure of social class to understand respondents' socioeconomic status.The climate strike data included a question asking respondents which social class they belong to (question wording: 'people sometimes describe themselves as belonging to the working class, the middle class, or the upper or lower class.Would you describe yourself as belonging to the. ..?' forming a five-state categorical variable.Due to the limited number of climate strikers who belonged to lower class, working class, and lower middle class, we grouped these three categories and compared them to those who belong to upper middle class and upper class (coded as 1).We refer to the latter category as high social class.
Lastly, we measured political interest with a dichotomous variable distinguishing between respondents who are interested in politics and those who are not.The question we asked was: 'how interested are you in politics?' with possible answer including 'not at all', 'not very', 'quite', and 'very'.We recoded the former two as 0 and the latter two as 1.

Regression models
We estimated linear probability models using OLS.We ran distinct models for both dependent variables.Firstly, we ran a model including our independent variables and our control variables to assess the influence of main predictors.The results are presented in graphs which show the effects of all independent variables and their 95% confidence intervals to facilitate the visualization and interpretation of results (the corresponding regression table is available in Appendix III).Due to the dichotomous nature of our dependent variables, we also ran logistic regression models as a robustness check (see Appendix IV).

Emotions
We begin our empirical analyses by comparing anger and fear among young and older climate strike demonstrators.We asked demonstrators to what extent they feel angry or fearful about climate change.Table 1 shows that roughly 40% of all respondents felt very angry about climate change.The percentage varied between 35.8% of adults and 45.2% of senior citizens.In addition, another third said that they feel quite angry.In this case, the percentages were very similar across age categories: they range between 30.7 and 29.0%.For those who did not feel angrer or very little , the percentage was smaller among senior citizens (only 24.5%) and the difference relative to the other two groups, youth and adults, was statistically significant.Among the youth and adults, 30.8% and 33.5%, respectively, did not feel angry about climate change.
The picture is different when investigating fear.In general, the share of respondents reporting being very fearful is smaller than the share of respondents being very angry.However, young people were more likely than the other age groups to be very fearful about climate change.22.9% of the youth felt very fearful, which compares to 15.5% of adults and 12.7% of senior citizens.The share of respondents who mentioned that they are quite fearful was more similar in the three age groups, with roughly a quarter of respondents mentioning feeling fearful (26.7% of the youth, 24.8% of adults, and 26.9% of seniors).Overall, young people are more likely to feel fearful than older generations about climate change and the differences between age groups are significant.

Motivation in joining the demonstration
Secondly, we analyzed two motivations to join the demonstration to understand the role that they play in different age groups.Table 2 shows that motivations varied according to age groups.While a larger share of young people mentioned that they participated to defend their own interests, among adults and senior citizens a larger share participated to pressure politicians.Regarding the defense of personal interests, this motivation was mentioned by 40.6% of the young demonstrators that we surveyed.This is compared to 34.6% of adults and only 28.8% of senior citizens.Adjusted residuals showed that the share of the youth who wanted to defend their interests was larger than expected.Conversely, among seniors it was smaller than expected.This means that there was  We use adjusted residuals to identify statistically significant differences between age groups.We use bold to indicated adjusted residuals >1.962, this means that the percentage of respondents in this category is higher than expected.We use underscore to indicate adjusted residuals < −1.962; in this case the percentage of respondents in this category is smaller than expected.
a statistically significant difference between these groups regarding the motivation to defend one's interests.Regarding the motivation to pressure politicians, a larger share of adults and senior citizens demonstrated to pressure politicians than among young people.Among the youth, 63.4% agreed or totally agreed with this motivation.Conversely, for adults, the percentage was 74.1%.For seniors, it was 77.5%.

Explaining motivations to protest among young, adult, and senior citizens
Since our dataset only contains demonstrators, we are unable to predict participation.However, we can use our data to better understand the motivations to participate among different age groups.To do so, we examined the predictors of the two main motivations to participate (self-interest and pressuring politicians).More specifically, we investigated the effects of age group categories and fear and anger on demonstrators' motivation to participate, controlling for gender, perceived social class, political interest, and the country in which the demonstration took place.
Figure 1 presents the coefficients of two OLS regressions (see Appendix III for the corresponding regression table and Appendix IV for the same models estimated using a logistic regression).The left-hand side of the graph predicts the motivation to defend one's interests and the right-hand side panel predicts the motivation to pressure politicians.Compared to being senior (the reference category in the analysis), being young increased the likelihood of wanting to defend one's interests and reduced the likelihood of wanting to pressure politicians.This may be related to the fact that young people are those who will pay the highest cost for climate change throughout their lives and will experience higher personal stakes in the issue.At the same time, senior and adult citizens were more likely to consider pressuring politicians as an important goal when compared with the youth.This may be due to the higher level of trust in political institutions in their generation.More generally, this analysis confirms that the age group differences in motivations that are documented in the descriptive analysis are maintained when controlling for gender, perceived social class, political interest, and the country in which the demonstration took place.When controlling for the different factors mentioned above, those who felt very angry were significantly more likely to mention both motivations compared to those who are not angry (the reference category).Similarly, those who were very fearful were significantly more likely to mention that they want to defend their interests and to pressure politicians.For both emotions and both motivations, the intermediary categories of 'quite angry' and 'quite fearful' were found between the two more extreme categories.These results demonstrate that the intensity of emotions is associated with the motivations of individuals to protest.
The effect of anger is significant because feeling very angry (as opposed to not angry) increased the probability of declaring that one demonstrates to defend one's own interest or to pressure politicians by about 0.2.The effect of fear was less significant.Feeling very fearful increased the probability of demonstrating to defend one's own interest or to pressure politicians by about 0.1.Both effects were substantial when compared to other variables in the model.For example, the effect of political interest, which is the control variable that most systematically affects motivations to join, was less significant.
The analysis further demonstrates that the motivations to join the protest varied by country.The German-speaking countries (Austria, Germany, and Switzerland) and the United States stand out as the countries in which the highest share of strikers declared that they were participating to defend their interests.Conversely, in Italy, the likelihood of participating to pressure politicians was the lowest.

Conclusion
In this paper, we have examined the extent to which fear and anger contribute to different motivations to demonstrate in climate strike demonstrations in 2019, which took place in eight countries in Europe, Australia, and the United States.Building on existing research, we argue that both anger and fear are associated with climate change and are conducive to collective political action.Fear and anger ensuing from climate change and the lack of state action to mitigate the effects or to adapt to its consequences support collective political action (Kleres & Wettergren, 2017;Stanley et al., 2021).However, existing research did not examine how different age groups or generations feel about climate change.It did not address whether the emotions triggered by climate change are same for young and older citizens.
We argue that older citizens are less likely to feel frightened by the consequences of climate change because they are less likely to live long enough to experience its most severe consequences.However, they are more likely to feel angry because they have placed a stronger belief and trust in institutional politics, which might lead to indignation when they realize that the government is not acting on this urgent issue.Our empirical findings confirm that older citizens are less likely to fear the consequences of climate change when compared to the youth.They also show that senior citizens are more likely to experience anger in relation to climate change when compared to adults, but not to younger citizens.
In addition, we compared two different motivations for joining the climate strike demonstrations: to defend one's interest and to put pressure on politicians.Firstly, we found that young demonstrators were more likely to mention defending one's interest as a motivation to join the demonstration than older participants.Secondly, we observed that older citizens were more likely to mention pressuring politicians as a motivation to participate than younger citizens.In our regression models, we used emotions to predict these two motivations.We found that independently of the age of respondents, higher levels of anger were associated with higher levels of willingness to defend one's interests and to pressure politicians when participating in a demonstration.The same can be said about the effect of fear on these motivations, although the magnitude of this effect is less significant.Therefore, these emotions, which are distributed differently across age groups, contribute to demonstrators' motivations to participate in the strike.Overall, the findings reported in this study provide valuable insights in understanding the climate strike movement's heterogeneity.
However, this research has several limitations and identifies the need for future research.Firstly, our study focuses on the role of emotions for two motivations to participate.However, it does not provide evidence on the role of emotions for participation itself because it only focuses on demonstrators.Future research could analyze how emotions about climate change differ among individuals who participate and those who do not participate in climate strike demonstrations, and how emotions may play a different role for disparate age groups.Furthermore, although we have focused on anger and fear, research on emotions and politics makes it clear that climate change also triggers other emotions, such as hope or guilt.It would therefore be interesting to study them as well.Finally, our focus on individual factors overshadows the potential differences across contexts.The analysis pools data from different cities and demonstrations.In future research, it would be worthwhile to analyze whether and how the role of age and emotions varies across political and geographical contexts.

Notes
1. https://www.oecd-ilibrary.org/sites/99acb105-en/index.html?itemId=/content/component/ 99acb105-en.2. Note that the survey included other potential motivations for participating in a demonstration, such as feeling a moral obligation to do so, because someone asked me to join, or to express my solidarity.We chose to focus specifically on the two items mentioned above because they are the most closely linked with the concrete policy goals expressed by the respondents.3. 2 Our choice to focus on perceived social class rather than education is linked with the fact that many of the young protesters are still undertaking their education.Therefore, educational attainments are poor control variables in a comparison between different age categories.Young people are more likely to have lower levels of education because they are still in the process of acquiring their qualifications.

Table 1 .
Demonstrators' feeling angry and fearful when thinking about climate change.

Table 2 .
Demonstrators' motivations for taking part in the climate strike demonstrations.

(ref.) Not angry (ref.) Not fearful (ref.) Control variables Country (ref. Belgium)
Figure 1.OLS Regressions using age categories and emotions to predict different motivations to demonstrate (Doi=Defend one's interests and PP=Pressure politicians).