Retrieved articles were diverse in methodology and operationalization of independent and dependent variables. Most articles utilized a single study design (n = 30); the most represented were cross-sectional studies (n = 17), followed by longitudinal studies (n = 9). There were also some experimental (n = 2) and qualitative (n = 3) studies. Only four articles used a multiple studies design, usually combining a cross-sectional study with an experimental or longitudinal study.

Most studies were done on North American or Western European samples (n = 27), with several exceptions. Three studies were conducted on samples from Hong Kong, one from India, one from Malaysia, one from Latin American countries (Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Rep., Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay, Venezuela), one from New Zealand, and one from Poland. Additionally, one study included American, Italian, and Iranian subsamples. Regarding ethnicity, most studies (n = 26) had unspecified samples or samples mainly containing (more than 80%) people from majority populations (depending on the country in which the study was conducted, but usually Whites). Six studies had more mixed samples, but still, around half of the respondents were from the majority populations. One study focused on an immigrant sample (from African, Asian, European, and South American countries of origin), one study focused on Black and Latinx students, and one study focused solely on Blacks. Finally, one study focused on a sample of LGBT + adolescents.

The articles’ most significant source of diversity was the type of civic engagement. The papers used different measures; some measured more than one type of engagement. Different types of civic engagement are usually clustered into broader categories, an approach also used in this review. Thus, activities from the selected papers were clustered into three major forms of civic engagement based on categorizations present in the existing literature (Barrett & Zani, 2015; van Deth, 2014). First is the protest action, which can be defined as an unconventional activity (i.e., outside the established political institutions) targeting social and political issues (such as protesting, signing a petition, or activism). Second is social engagement, representing socially oriented activities focused on problem-solving and helping others (including donations, volunteering, or community participation). The last is conventional engagement, representing voluntary activities in the sphere of established political institutions (e.g., electoral activities, donating money to a political party). This final form can also include voting, although voting is often studied separately. The most frequently represented were activities falling into the category of protest action (n = 17), followed by social engagement (n = 13). Fewer studies were interested in conventional engagement (n = 9); most of them focused specifically on voting (n = 5). Apart from these major categories, there were two cases where authors were interested in organizational membership and two cases where authors did not specify or distinguish different types of engagement. While some authors (Kim & Hoewe, 2020; Oser et al., 2013) argued for distinguishing between online and offline activities, only five articles explicitly focusing on online engagement were found.

Regarding well-being as a dependent variable, most articles fell into one of the four main categories (some of them focused on more than one type of well-being). Most represented was emotional well-being (sometimes called positive affect or life satisfaction; n = 14), followed by mental health (n = 11), psychological well-being (n = 10), and social well-being (n = 9). Four studies measured well-being as a composite of emotional, psychological, and social well-being, so they could not investigate whether civic engagement affected these dimensions of well-being differently. Additionally, there were three cases where the measurement of well-being did not fit into the mentioned categories.

Because the primary focus of this review is on the distinction between different types of well-being, the following text is structured by this outcome variable. Thus, the results are clustered into general (i.e., non-differentiated), psychological, social, and emotional well-being. Additionally, research operationalizing well-being as mental health symptoms is reported in a separate part. Few papers used well-being indicators that did not fit into the mentioned categories. Some articles are mentioned more than once since they consider multiple types of well-being. Table 1 shows detailed information on the included studies.

Some studies focused on a composite of psychological, social, and emotional well-being, thus not making inferences about specific dimensions of well-being. These studies all measured well-being by the MHC-SF (Lamers et al., 2011). A study by Hayhurst et al. (2019) from New Zealand showed that social engagement (e.g., volunteering) positively affected well-being, whether it was a past engagement or an intention to become engaged in the future. Nicotera et al. (2015) corroborated this on American undergraduate students, referring to the well-being composite as flourishing, and found that community engagement was positively associated with it. Fenn et al. (2023) also supported the positive effect of social engagement on American non-college youth, showing that self-efficacy positively mediated the effect. However, another study on American undergraduate students (Fenn et al., 2021) showed no direct effect of social engagement (e.g., community engagement, volunteering), although its effect might have been entirely mediated by self-efficacy, as community engagement and volunteering were positively associated with self-efficacy, which, in turn, was positively associated with well-being. Overall, the most represented activities related to general well-being fell into the social engagement category and tended to positively affect well-being.

Regarding conventional engagement, Fenn et al. (2021) found that electoral activities (e.g., voting) positively affected well-being. These results were corroborated by Fenn et al. (2023), where the association was also positive. In both these cases, these effects were positively mediated by self-efficacy. This is similar to social engagement, as conventional engagement seems to also have a beneficial effect.

As for the other categories of civic engagement, the studies by Fenn et al. (2021, 2023) also explored protest action and found negative associations with well-being. Furthermore, Fenn et al. (2023) also found that this effect was negatively mediated by meaning in life. The authors were also interested in online engagement, which was also negatively associated with well-being (Fenn et al., 2023). Contrary to the previous forms of engagement, protest action and online engagement seemed to have rather detrimental effects. Generally, due to the combined measure of well-being, the possible different effects on specific well-being dimensions could not be examined based on the mentioned results.

Cross-sectional studies focusing on psychological well-being showed that it was positively affected by different forms of civic engagement, such as online collective action (online engagement; Foster, 2014, 2015, 2019), conventional activism (conventional engagement; Klar & Kasser, 2009), human rights activism (protest action; Montague & Eiroa-Orosa, 2017), and prosocial activities (social engagement; Chan & Mak, 2020). However, in a study by Klar and Kasser (2009) on American youth, protest action specified as high-risk activism (e.g., activities with a higher risk of arrest), was not associated with psychological well-being. This corroborates with the cross-sectional research of Costabile et al. (2021) on Italian adolescents, where authors tested the opposite direction of effects and found that psychological well-being was positively associated with activism (i.e., political action through nonviolent means) but negatively with radicalism (i.e., political action through violent means). Their results also showed that these associations were mediated by social disconnectedness and the perceived illegitimacy of the authorities (Costabile et al., 2021). The qualitative study of Montague and Eiroa-Orosa (2017) on adolescents from the United Kingdom identified that the beneficial effect of protest action stemmed mainly from being active (rather than passive) and belonging to a larger group represented by a political movement. Generally, results from these studies show rather beneficial effects of civic engagement.

Foster (2014, 2015) was interested in the role of intervening variables in the case of online engagement and found that, at least among undergraduate women from Canada, the effect of online collective action depended on the perceived pervasiveness of discrimination (sexism). The beneficial effects only occurred when the perceived pervasiveness was high, but the effect was detrimental when there was no perceived pervasiveness (Foster, 2014, 2015). Additionally, Foster (2015) showed that another form of online collective action (tweeting) positively affected psychological well-being. However, in her subsequent study on a similar Canadian sample, Foster (2019) found that online activism was beneficial only in interaction with gender identity, meaning that engagement benefits well-being only when a person has a strong social identity, which serves as a protective factor. Overall, it seems that the effect of online engagement is more mixed and dependent on other factors.

However, it should be kept in mind that the results mentioned above come from cross-sectional studies, and thus, it can be premature to infer causality from them automatically. A longitudinal study on college students from Hong Kong by Chan and Mak (2020) showed that social engagement did not predict psychological well-being. Another longitudinal study on Hong Kong youth by Chan et al. (2021) even showed that more avid and long-term engagement in social movement (which would fall into the protest action category) had a detrimental effect on psychological well-being. Contrary to the cross-sectional and experimental results above, longitudinal results do not support the beneficial effect of civic engagement.

Social well-being is mainly studied in relation to two forms of civic engagement: social engagement and protest action. Regarding social engagement, existing literature showed that helping activities such as volunteering, donating to charities, or organizing events have primarily beneficial effects on social well-being among the youth from Italy (Albanesi et al., 2007; Cicognani et al., 2008), Hong Kong (Chan & Mak, 2020), and various European countries (Pavlova & Lühr, 2023). This was also corroborated by Chan and Mak’s (2020) longitudinal study, in which social well-being was the only dimension positively affected by prosocial activities. Although not precisely social engagement, Albanesi et al. (2007) and Cicognani et al. (2015) showed that, among the Italian youth, membership in prosocial organizations, including charity or youth groups, was also beneficial in this matter. In sum, the positive effect of social engagement on social well-being can be considered robust, supported by both cross-sectional (Albanesi et al., 2007; Cicognani et al., 2008; Cicognani et al., 2015; Chan & Mak, 2020; Pavlova & Lühr, 2023) and longitudinal (Chan & Mak, 2020) research.

The evidence in the case of protest action is mixed. Sohi and Singh (2015) conducted a study on Indian Northeasterners residing in Delhi and found that being engaged in protest action had a positive effect on social well-being, attenuating the otherwise negative effects of perceived discrimination. However, Albanesi et al. (2007) showed that protest action was unrelated to social well-being among Italian adolescents. In contrast, the longitudinal study on Hong Kong youth by Chan et al. (2021) showed that avid protest action and online engagement had a detrimental effect. These contradictory results might point to the role of perceived costs and riskiness of protest action. Sohi and Singh (2015) investigated engagement due to an acute perception of discrimination, whereas Albanesi et al. (2007) explored general participation in various protest activities during the previous year. Chan et al. (2021) studied a more long-term engagement in a social movement with a potentially increased risk of burnout and, thus, deteriorated social well-being. Thus, the effects of protest action are more mixed, similar to the case of psychological well-being mentioned above.

Compared to psychological and social well-being, the effects on emotional well-being are much more inconclusive. Because emotional well-being represents current emotional experiences and thus is related to various affective states, studies that measured this dimension of well-being as positive affect, happiness, and satisfaction with life, which are known indicators of emotional well-being (Keyes, 2002), were also included. Klar and Kasser (2009) found that, among American youth, low-risk activism (corresponding with a conventional form of engagement) was positively linked to life satisfaction and positive affect. Weitz-Shapiro and Winters (2011) corroborated this on a sample from various Latin American countries. However, Lorenzini (2015) tested these associations among employed and unemployed Swiss youth and found that life satisfaction was negatively associated with contacting political institutions, but only among employed youth. These results can suggest that conventional engagement is rather beneficial for emotional well-being.

In the case of protest action, Becker et al. (2011) found that collective action was positively associated with positive affect among German youth. Also, Lafreniere et al. (2023) found that Black Lives Matter (BLM) activism (mixed protesting and online participation) among Black young adults living in Canada had a positive cross-sectional association with life satisfaction. This was also corroborated by Klar and Kasser’s (2009) results on high-risk activism (representing protest action). However, a longitudinal study by Chan et al. (2021) showed no effect on emotional well-being in a sample of Hong Kong youth. Lorenzini (2015) found that life satisfaction was positively associated with protest action only among unemployed Swiss youth, whereas there was no supported association among employed youth. In sum, only cross-sectional results support the beneficial effect of protest action.

Regarding other forms of civic engagement, some studies showed no relation between social engagement and emotional well-being among Hong Kong youth (Chan & Mak, 2020), Canadian youth (Fang et al., 2017), and Malaysian youth (Zaremohzzabieh et al., 2019). Foster (2015) found that even though online engagement led to decreased negative affect (sadness), it was unrelated to positive affect (happiness), at least among Canadian undergraduate women. In a similar Canadian sample, online engagement had a beneficial effect on life satisfaction if this engagement was perceived as active and not passive, and this effect was moderated by gender identity (Foster, 2019). Thus, it seems that other forms of civic engagement can have only limited beneficial effects.

Some studies also addressed the question of causality. Weitz-Shapiro and Winters (2011) used cross-sectional comparisons based on eligibility to vote to identify causal relations and found a positive association between voting and life satisfaction. However, their further analysis did not provide evidence that voting (conventional engagement) increased life satisfaction, but that satisfied people were more likely to vote. The longitudinal study of Fang et al. (2017) also supported this causal direction because, in their model, happiness predicted social engagement and fit the data better than the model with opposite causality or bidirectional effects. On the other hand, the longitudinal study of Lafreniere et al. (2023) supported the opposite direction, that protest action positively predicted life satisfaction. This effect might be explained by the fact that this study was conducted among Black young adults, thus representing a marginalized group experiencing discrimination and oppression. However, their study tested only for this particular direction and thus could not infer possible reciprocal effects. Overall, it seems that, in this case, the effects are reversed, as emotional well-being can be considered a predictor of civic engagement rather than an outcome.

Some studies understood well-being as mental health, defined as the absence of depressive symptoms, psychosomatic symptoms, or anxiety. The most often studied indicator was depressive symptoms. Existing literature showed that different forms of civic engagement, such as social engagement among American youth (Ballard et al., 2019; Hope et al., 2018; Kim & Morgül, 2017; Wray-Lake et al., 2019), conventional engagement among American adolescents (Ballard et al., 2019; Wray-Lake et al., 2019), and online engagement among Canadian undergraduate women (Foster, 2014) led to a decrease in these symptoms and thus could be considered beneficial for mental health. On the other hand, Wray-Lake et al. (2019) supported conventional engagement (e.g., contributing money to a political party) as a positive predictor of depressive symptoms in a sample of American youth, but this was not true for voting, which was measured outside other conventional activities. Furthermore, online participation was associated with increased depressive symptoms among American LGBT adolescents, and this relationship was fully mediated by exposure to web-based discrimination (Tao & Fisher, 2023). Regarding protest action, Ballard et al. (2019) found no relation to depressive symptoms, while Hope et al. (2018) showed that these activities can mitigate the negative effects of racial/ethnic discrimination among Latinx students, thus having a beneficial effect. However, Hope et al. (2018) also showed that among Black students, the effect of protest action was the opposite, meaning that it could exacerbate the experiences of discrimination among these students. Schwartz et al. (2023) also showed mixed results on a sample of American college students. While protest action (i.e., climate activism) was overall not directly associated with depressive or anxiety symptoms, it had some b