Introduction
The news media environment and the ways citizens engage with politics have profoundly changed in the past decades. The relationship between news consumption and political participation has become more complex. We witness several interconnected trends that make it more difficult for citizens to engage with news, creating a more pressing need to develop news media literacy to foster a better-functioning democracy. These trends encompass a changed news media environment, the proliferation of dis- and misinformation, and diminishing trust in professional media and journalism (van Aelst et al., 2017;Vraga and Tully, 2021;Hanitzsch, Van Dalen and Steindl, 2018).
First, the contemporary news media environment could be described as a complex, high-choice political information environment (Van Aelst et al., 2017), where audiences have an opportunity to choose among a myriad of news sources on many platforms. This might be challenging for news consumers with less media literacy. Second, the proliferation of disinformation and misinformation in the last decade signifies an additional challenge for news consumers. In the recent period, the academic research focus has greatly shifted toward analysing misinformation and its effects on political processes and democracy. One strand of research also focuses on the role of news media literacy in combating misinformation, especially on social media (Vraga and Tully, 2021). Third, trust in professional media and journalism is eroding in many countries, which is connected to the decline of trust in other social institutions as well (although this decline is not universal, seeHanitzsch, Van Dalen and Steindl, 2018).
These trends are the focus of many studies in political communication, as they are believed to contribute to news avoidance and the decline in political participation. News media literacy (NML) could, in this way, be seen as a remedy to declining levels of trust in news, news avoidance and disengagement with political and civic life. However, the findings about the role of NML in political participation have been ambiguous, and there have so far been no such studies in Croatia. The goal of this study is to explore the relationship between NML and various dnsions of political participation by considering generational differences between news audiences.
The Importance of News for Political Participation
A positive relationship between news consumption and political participation has been long established in political communication research (see an overview of studies inVozab, 2019a). There are, however, some differences between various media in their effects. Most studies found newspapers and the press to be in a “virtuous circle” with political participation, in the sense that those who are engaged use newspapers more often, which then feeds into more participation (Norris, 2000). However, newspapers are usually read by audiences of higher socio-economic status and there is a large gap in political knowledge and participation between newspaper readers and non-readers. Television is, in this sense, an equaliser, but its impact on political participation is often not significant, or it somets has a “media malaise” effect that demobilises citizens (Vozab, 2019a). A meta-study about the relationship between digital media and political participation spanning 20 years and 50 countries found that the effects of digital media on political participation are increasing with t, meaning that digital media are becoming increasingly important for contemporary political participation (Boulianne, 2020).
In Croatia, the relationship between news consumption and political participation has been researched by several authors. Lamza Posavec and Rihtar (2003) found that press audiences were much more likely to vote in elections, which confirms an established correlation between newspaper reading and political participation. Vozab (2019a) found that news consumption, specifically reading newspapers and having wider news repertoires, mobilised citizens only to participate digitally. On the other hand, citizens who followed commercial television more often were less likely to engage in demonstrations and digital participation (Vozab, 2019a).
Extending Forms of Political Participation1
Political participation has been conceptualised and measured in different ways over the past 60 years since the pioneering research of civic culture by Almond and Verba (1963). They conceptualised political participation in its institutional, traditional, or conventional form, such as voting, supporting political campaigns, or contacting professional politicians and the government. This type of institutional participation is elite-oriented and encompasses the engagement of citizens with the system of representative politics (Teorell, 2006). Over time, repertoires of political participation have expanded, either because of shifts in the political system and expanded opportunities for political action, or because of social changes and the development of digital technologies. According to Dalton (2008), the traditional norms of citizenship associated with institutional participation are losing their importance with the process of (post)modernisation. They are increasingly being supplemented by norms of engaged citizenship and extra-institutional or unconventional forms of political participation. Norris (2002) argued that the operationalisation of political participation also needs to include its extra-institutional or unconventional dnsion, as she defined political participation in a much broader sense, as all the behaviours which are a part of civil society, ad at effecting systemic change in political or public issues. These practices which include protests, participation in new social movements, petitions, occupations, and political consumerism, have become more common and normalised since the 1960s, no longer considered “unconventional” (Theocharis and van Deth, 2019: 18).
The conceptualisation of political participation has become more complex with the advent of digital technologies and their increasing intertwining with political practices. While digital media and technologies were initially perceived as merely tools for political participation, in the recent period they are seen as technologies that helped shape completely new forms of political participation (Theocharis, 2015, Theocharis and van Deth, 2019). However, expanding the conceptualisation of political participation has also created issues by making the concept overly broad and less clear (Theocharis and van Deth, 2019). This is especially the case with digital political practices, often criticised as being nothing more than slacktivism, not having real political importance. To resolve this issue, Theocharis and van Deth (2019) provide criteria for defining which digital practices could belong to a dnsion of political participation.
Theocharis and van Deth (2019), define political participation based on the following conditions: a) it is an activity b) it is carried out by a citizen (not a professional politician) c) it is voluntary and d) it focuses on the government, politics or the state in a broader sense and e) it includes issues related to collective action, community or the public good and f) it is politically motivated or targeted, e.g. it is ad at putting a political issue on the agenda. This wide definition of political participation encompasses diverse repertoires, ranging from traditional forms of participation (voting or supporting a political campaign), to expressive forms that often take place in the digital sphere (tweeting about political issues or boycotting products for political reasons).
The digital realm is currently where new forms of participation proliferate the most. Theocharis (2015: 6) labels such participation as digitally networked participation (DNP) and defines it as “a networked media-based personalized action that is carried out by individual citizens with the intent to display their own mobilization and activate their social networks in order to raise awareness about, or exert social and political pressures for the solution of, a social or political problem”.
Generational or Life-Cycle Perspective on Political Participation and News Consumption
The dominant age-related approaches to political participation are the life cycle and generational perspective. For example, Norris (2003) found life-cycle effects in citizenship norms and institutional political participation. As younger age cohorts enter middle age, they tend to become more interested in politics, are more inclined to identify themselves with political ideologies or political parties, and are more ready to vote and participate in civil society. This pattern particularly stands for the institutional forms of political participation, correlating with older age groups, while younger age cohorts might be more interested in extra-institutional forms of political participation, associated with the norms of engaged citizenship (Norris, 2003).
Most analyses from the generational perspective rely on the work of Karl Mannheim (2015). He argued that not only social class but also generation is a social unit that brings social change. Generation is defined as people who were born in s similar place and period and shared similar experiences throughout their formative years. Such experiences, which form a specific generational Zeitgeist, are historical events, such as World War II, the Moon landing, the Chernobyl disaster, and the fall of the Berlin wall. According to Mannheim (2015), the “formative years” during which important experiences occur are between 15 to 25 years of age. Grasso (142016) assigned respondents to a generation if they spent the majority of their formative years in the historical period defining that generation.
According to the generational argument, the levels of political participation are connected to the socialisation process specific to a certain generation. Therefore, we could speak of generations that are more “participative” (e.g., the generation of students during 1968), and those that are more passive. Some authors associate unconventional forms of political participation with generational changes (Bennett, Wells and Rank, 2009). Through the process of modernisation, the quality of life and education levels are rising, provoking the development of postmaterialist values, the greater importance of autonomy and self-actualisation (Inglehart and Welzel, 2007), and norms of engaged citizenship (Dalton, 2008). Following Dalton, Bennett, Wells and Rank (2009) suggest that generational changes produce two main types of citizens. Dutiful citizens have a strong sense of responsibility to vote, are more inclined to institutional forms of participation, and exhibit a higher level of trust in political parties, the government, and traditional media. On the other hand, self-actualising citizens have a weaker sense of duty to participate and are more engaged in “lifestyle politics” (volunteering, activism), have less trust in political parties, the government, and traditional media and are more inclined to consume news through digital media.
If the generational argument is correct, the levels of participation should have a certain stability within a generation regardless of the life-cycle, and also have the potential to bring social change (Grasso, 2016). Therefore, other authors suggest that extra-institutional or unconventional forms of political participation could be a generational trend. For example, when “protest generations” get older they might still be more active than younger generations not socialised in the same political environment. For example, Grasso (2016: 39) argues that generations socialised in the 1980s and 1990s might be less participative than 1960s generations, despite the rising levels of education. She explains this by the rising influence of neoliberal individualism and the end of the Cold War, which diminished the strong ideological struggles that defined previous generations in Europe. Some comparative research shows that younger age cohorts are less inclined to participate through traditional forms of participation, but are more active through unconventional practices, which would support the life-cycle hypothesis (Marien, Hooghe and Quintelier, 2010). However, another generational analysis found that each younger generation does not linearly engage more in unconventional forms of participation. On the contrary, the “baby boom” generation, socialised in the period of radical politics in Western Europe in the 1960s, still engages more in protests, regardless of ageing (Grasso, 2016: 164). The generational perspective offers research opportunities to evaluate how the relationship between news consumption and political participation might look in the future (Andersen et al., 2021).
The generational argument is also relevant for explaining political participation in post-socialist societies. Levels of political participation are lower in post-socialist countries due to the socialisation of socialist generations in authoritarian political systems where democratic political practices could not develop properly (Hooghe and Quintelier, 2014). Another perspective argues that most post-socialist countries are also burdened with higher levels of corruption, a lack of good governance and economic downturns, which demotivate citizens who experienced the transition from political participation (Hooghe and Quintelier, 2014). Hooghe and Quintelier, (2014) show that both explanations are correct, with older age cohorts in post-socialist countries participating less.
In studies of age-related differences in news and media consumption, authors have usually studied the effects of various age cohorts rather than generations. Many of the studies also use a well-known and widely used categorisation of generations: the Baby Boom generation, Generation X, Millennials and Generation Z (seeAndersen etc. al., 2021;Diehl, Barnidge, and Gil de Zúñiga, 2019;Vozab, 2019b;Zukin et al., 2006). However, this categorisation is criticised because it does not follow the “Mannheimian” definition of generations, it is seen as simplistic, ahistorical and fails to consider the political and cultural contexts that defined different generations (Bolin, 2017). Rather than a generational analysis, these studies perform an analysis of differences between different age cohorts. Media generations, on the other hand, could be defined as those whose shared experience is connected to their “fresh contact” with a new media technology (Bolin, 2017). In Croatia, media generations include the generation of traditional media users, the generation of digital immigrants, and the youngest generation, digital newspaper seekers and digital newspaper avoiders (Čuvalo and Peruško, 2017).
Blekesaune, Elvestad and Aalberg, (2012) found that younger age cohorts are more likely to “tune out” and consume traditional news less often. Andersen et al. (2021) propose an Engagement-Participation-Information-Generations model to study generational effects in the relationship between news consumption and political participation. They find that news consumption rises linearly with age, that social media is used much more by younger generations, and that the levels of political participation are stable across generations (Andersen et al., 2021). The authors also found that younger generations benefit more from increased news exposure in terms of political participation, and that they participate more in shorter-term political activities (Andersen et al., 2021). Diehl, Barnidge, and Gil de Zúñiga (2019) show that there is no significant difference between different age cohorts in overall news use. However, they find a difference in the way news is consumed: younger generations tend to be “news snackers”, as “information is consumed in smaller amounts through a variety of technical modalities” (16). They also demonstrate that multi-platform news consumption has a stronger impact on political participation among younger age cohorts in comparison to older age groups (Diehl, Barnidge, and Gil de Zúñiga, 2019). Vozab (2019b) found that millennials differ from other generations by using fewer news media sources and relying more heavily on social media and commercial television as sources of news. They are somewhat less interested in news, trust media less and are more concerned about misinformation.
Media Literacy, News Media Literacy and Relationship with Political Participation
Building skills in different dnsions of media literacy is considered a key factor for engaged citizenship (Mihailidis and Thevenin, 2013). Some media literacy competencies are associated with the basic tenets of engaged citizenship. Mihailidis and Thevenin (2013) highlight these characteristics of media-literate active citizens: critical thinkers, creators and communicators, and agents of social change. Critical thinking skills in reading media messages help citizens gather reliable and relevant information about their society and question authorities (Mihailidis and Thevenin, 2013). This approach is inspired by Paulo Freire’s critical pedagogy, where civic engagement is associated with critical thinking (Mihailidis and Thevenin, 2013). Creation and communication skills are associated with the norms of self-actualising and engaged citizens (Mihailidis and Thevenin, 2013), and are related not only to information consumption but also to using technology for content creation and communication. These skills seem to resemble digitally networked participation as discussed by Theocharis and van Deth (2019). Finally, according to Mihailidis and Thevenin (2013), media literacy should promote citizen skills that bring social change and contribute to the political sphere by creating new initiatives.
News media literacy (NML) is a subset of media literacy focusing on the skills and knowledge needed for the critical assessment of news media (Ashley, Maksl and Craft, 2013). To become news literate, one must acquire “knowledge and skills regarding news content, systems, and practices”, thus understanding not only the media texts they consume but also how they are produced in media industries and the effects they have on citizens (Vraga et al., 2021: 2). News literate citizens should also understand the normative role media have in a democracy, and the structural constraints that affect the ability to exercise that role (Ashley, Maksl and Craft, 2013: 8). Ashley, Maksl and Craft (2013) developed an attitudinal NML scale which consists of three domains: authors and audiences, messages and meanings, and representation and reality. The goal of the scale is to assess the understanding of how media (authors) target audiences, how messages can have different meanings for diverse audiences, and how media can frame issues (Ashley, Maksl and Craft, 2013:11). NML is associated with various behavioural outcomes and other skills, such as “news consumption, identification of misinformation, rejection of conspiracy beliefs, and civic engagement” (Vraga et al., 2021) or scepticism about the information on social media (Vraga and Tully, 2021).
The empirical analyses of the impact of media literacy on political participation are on the rise, somets yielding mixed results. Mihailidis (2009) analysed the effects of media literacy programmes on participation and attitudes towards politics. His study found that media literacy programmes empowered students to critically assess media messages across different media formats (Mihailidis, 2009). However, students who attended media literacy programmes also expressed negative and cynical opinions about the role of media in society (Mihailidis, 2009). A panel study on the impact of media literacy programmes on the political engagement of high school and university students in the US found that these programmes had a positive association with political engagement and exposure to diverse viewpoints (Kahne, Lee and Feezell, 2012). A similar study also demonstrated that exposing youth to “digital engagement learning opportunities” had a positive outcome for online participation and targeted political pressure (Kahne and Bowyer, 2019). Tully and Vraga (2018), in their mixed-methods study, found that NML messages promoted self-confidence in the critical assessment of news, which correlated with increased political efficacy. However, some of the respondents who perceived themselves as news literate were more sceptical of the media and political system and therefore felt less politically efficacious (Tully and Vraga, 2018: 781). Ashley et al. (2017) found a positive correlation between NML, political knowledge and political efficacy, but also discovered a negative relationship between NML and trust in the political system. They also found that NML did not correlate with political participation and identified a large news media literacy gap among audiences. Vraga and Tully (2021) found that NML reduced exposure to and sharing of political information on social media. Therefore, while NML could promote a critical understanding of news production and content, along with feelings of self-efficacy, somets it is unclear whether it fosters positive scepticism towards information (as it was shown byVraga and Tully, 2021), or cynicism and mistrust, potentially discouraging citizen engagement.
The primary means of acquiring NML is through the educational system. In Croatia, media literacy is not promoted and taught systematically but rather through the engagement of various stakeholders, mostly from civil society (Kanižaj, 2016). A media literacy programme has been introduced into formal education as “media culture” within the curricula of the Croatian language and culture. The programme has been evaluated as outdated, with students exhibiting high and diverse media use, but weaker critical thinking skills (Ciboci, 2018). On the other hand, news or civic media literacy, ad at building competencies for engaged citizenship in participatory democracy (Mihailidis and Thevenin, 2013), has been promoted only through informal education programmes. In the context of media literacy research in Croatia, the focus has predominantly been on children and youth. Until now, there have been no analyses of NML. This study, therefore, aims to contribute by analysing NML and its role in civic life, extending its scope beyond youth to different generations in Croatia.
Research Questions and Hypotheses
This study uses the news media literacy scale (Ashley et al., 2013) and the extended political participation scale (Theocharis and van Deth, 2019) to analyse the relationship between NML and various dnsions of political participation, considering generational differences among news audiences.
RQ1: What is the relationship between NML and various dnsions of political participation?
H1: Respondents with a higher score on the NML scale will engage more in institutional participation.
H2: Respondents with a higher score on the NML scale will engage more in unconventional types of participation.
H3: Participation in media literacy education programmes will correlate positively with institutional participation.
H4: Participation in media literacy education programmes will correlate positively with unconventional types of participation.
RQ2: What is the relationship between belonging to different generations and NML?
H5: Generations will significantly differ in their score on the NML scale.
RQ3: What is the relationship between generational belonging and political participation?
H6: Older generations will be more inclined to institutional participation.
H7: Younger generations will be more inclined to unconventional types of participation.
Methodology, Data and Sample
The analysis is based on data gathered through a quantitative study conducted on a quota sample representative of the population of Croatian citizens aged 18 and older (N=1033). Data was gathered by a combination of an online panel (84% of the sample) and face-to-face surveys (16%).2 Quotas were defined based on region, gender, age, and the size of the settlement. Data collection took place from 28 February until 15 March 2022. The quantitative study was conducted as part of the project "Media education is important (MOV)", led by the civil society organisation GONG, with the Institute for Development and International Relations in Zagreb as the main academic partner. The questionnaire was developed by the project research team, which included the translation to Croatian of items that were already developed and tested by other authors (e.g., the news literacy scale).
Measurement and Scales
Dependent variables
Political participation. The battery of questions on political participation was taken from the study by Theocharis and van Deth (2019: 121). The battery consists of fifteen items pertaining to various dnsions of political participation, including engagement with political parties, volunteering, and demonstrations, and participation in the digital sphere. Factor analysis was employed to discern different dnsions of political participation. With Principal Component Analysis as the extraction method and Varimax rotation with Kaiser Normalization, five factors were extracted, which together explain 61.93% of the variance. The factor solution partly resembles the one found in the study by Theocharis and van Deth (2019). The first factor consists of four variables: Volunteered in a social or humanitarian organisation, Volunteered for a community project, Worked or volunteered for a civic action group, and Donated money to a social or humanitarian organisation. Theocharis and van Deth (2019) labelled a similar factor “volunteering”, and in this study, the factor was labelled “civic engagement”. Based on this factor, a composite variable was created, consisting of three variables with the highest saturation and consistency (m=5.61, sd=0.84, Cronbach Alpha=0.793). Based on the extracted factors, another three composite variables pertaining to dnsions of political participation were created: institutional participation (m=3.86, sd=0.46, Cronbach Alpha=0.75), DNP (m= 3.62, sd=0.71, Cronbach Alpha=0.77), and political consumerism (m=3.64, sd=0.65, Cronbach Alpha=0.62).3
Independent variables
News media literacy. The NML scale was adopted from Ashley, Maksl and Craft (2013), who developed and tested the scale across various samples. The scale consists of 13 items.4 Ashley, Maksl and Craft (2013) found that, across samples, factor analysis of the variables in the scale yielded a one-factor solution with high internal consistency. Additionally, the scale correlated with current events and media system knowledge. For this study, factor analysis was performed on the NML scale variables, which also resulted in a one-factor solution, consistent with the findings of Ashley, Maksl and Craft (2013), with high internal consistency (Cronbach Alpha=0.85). Similar to the findings in Ashley, Maksl and Craft (2013), respondents scored highly on the scale (the item “news about conflicts receive more attention” had the highest mean m=4.45, sd=0.72 while the item “audiences seek news which corresponds to their political views” had the lowest mean (m=3.84, sd=0.91). An additive index was created from all the items in the NML scale.
Media literacy education. Respondents were asked if they ever participated in media literacy education programmes (within the formal education system, through workshops, etc.). Only 10.2% of the respondents in the sample participated in such programmes. The question that followed was open-ended, inviting respondents to explain what kind of programmes they participated in. Many of them mentioned either formal education programmes at schools or universities, and many also mentioned workshops, mostly relating to information and communication technology (ICT).
Pluralist media orientation. A composite variable (m=14,68, sd=2,94, Cronbach Alpha=0,771) was created from the following items: Media should report about national minorities; Media should report about gender equality; Media should report about people with different sexual orientations; Community media are important for democracy. These items were measured using a five-point agreement Likert scale. The variable measures attitudes towards the representation of diversity in media and towards media pluralism. This variable might correspond to post-materialist values that usually correlate with unconventional forms of participation. On the other hand, it would be interesting to see if the variable correlated with the NML scale.
Perception of misinformation online. Respondents were invited to answer the question “Somets it is difficult for me to evaluate if the information I found on the internet is true or false”, and to agree with the statement using a five-point Likert scale. This variable (m=3.11, sd= 0.92) measures whether citizens feel self-efficacy in recognising misinformation online.
Media consumption. Respondents were asked to evaluate their frequency of television, radio, press and internet use. Television, radio, and press consumption were measured using a four-point scale (from “never” to “every day or almost every day”). Internet use was measured using a five-point scale (from “less than once a month” to “every day”).
Generational groups. Generations were conceptualised based on Mannheim’s definition, positing that generations are formed by important shared experiences during formative years. As suggested by Grasso (2016), it is best to assign a respondent to a generation if he/she spent most of their formative years in a certain historical period. Therefore, the generations were defined as follows: Postwar generation (born between 1945 and 1954), Socialist crisis generation (born between 1955 and 1966), War generation (born between 1967 and 1979), Transitional generation (born between 1980 and 1990) and Post-transitional generation (born from 1991 onwards). The postwar generation, during their formative years, experienced a liberalisation of the socialist system after the introduction of the self-management model. They were also influenced by the student protests of 1968 and the protests in Croatia in 1971 (MASPOK). In terms of media use, this generation is mostly characterised as “traditionalist”, relying mostly on television and other traditional media (Čuvalo and Peruško, 2017). The socialist crisis generation, during their formative years, experienced multiple crises: the economic downturn following the adoption of the 1974 Constitution, the oil crisis in 1979, the death of Tito, the Kosovo protests and the growth of nationalist ideas. This generation is characterised by diverse media repertoires. Some members are traditionalists, some are digital immigrants, and some are digital users (Čuvalo and Peruško, 2017). The war generation witnessed Milošević’s rise to power, the fall of the Berlin Wall, the breakup of Yugoslavia, the war, and the immediate aftermath of the war. This generation is less associated with traditionalist media repertoires and could be described as digital immigrants and digital users (Čuvalo and Peruško, 2017). The transitional generation lived their formative years after the war and witnessed a period of transition from the semi-authoritarian system under Tuđman to a more functional democracy progressing towards EU membership. Finally, the post-transitional generation witnessed the consequences of the financial crisis of 2008, the EU accession, and the rise of populism and the radical right. This generation could be labelled as the “disillusioned generation”, a term coined by Ilišin and Spajić-Vrkaš (2017). Both the transitional and post-transitional generations could be described mostly as digital when it comes to their media repertoires (Čuvalo and Peruško, 2017).
Left–right placement. Respondents were asked to assess their own left-right placement on a ten-point scale (with a value of 1 indicating extreme left and a value of 10 indicating extreme right; m=5.16, sd=1.97).
Finally, socio-demographic variables such as income, education, and gender are included in the regression model.
Analysis
To explain differences among generational groups, the Brown-Forsythe Test was used.5 Hierarchical regression analysis was used to explain distinct types of political participation: institutional participation, civic engagement, DNP, and political consumerism. It was built with three blocks of independent variables. The first block consisted of socio-demographic variables (gender, education, and income), left–right placement and belonging to different generational groups, with the postwar generation serving as a reference category. The second block consisted of variables referring to media use, while the third block consisted of variables regarding participation in media literacy programmes, NML, pluralist media orientation, and perception of misinformation. The analysis was performed using the SPSS statistical analysis software.
Results
Differences in political participation among generations are shown in Figure 1. In almost all the types of participation, we can see a linear trend of rising levels of participation with younger generations. The least engaged generation is the postwar generation, with especially low levels of unconventional types of participation (civic engagement, m=0.23, sd=0.76; DNP, m=0.13, sd=0.46; political consumerism, m=0.18, sd=0.49). All generations exhibit the lowest levels of engagement when it comes to institutional participation, and the gap between institutional and unconventional participation rises with younger generations. However, the gap does not mean that there is a decline in institutional participation with younger generations – on average, they participate more in both institutional and unconventional forms.
The Brown-Forsythe (B-F) test found that generations differ significantly in DNP (F (4. 955)=6.84, p<.001) and political consumerism (F (4, 926)=4.47, p<.01). The Games Howell post-hoc test showed that the post-war generation exhibits a significantly lower DNP than the socialist-crisis (p<0.05) and all other generations (p<.001). The post-war generation also engages significantly less in political consumerism compared to the war (p<0.05), transitional (p<.001) and post-transitional (p<.01) generations. As there is no significant difference among the generations in institutional participation, H6 is rejected. These are also types of participation that exhibit the clearest linear positive trend with younger generations, although they are in a slight decline among the youngest, post-transitional generation. Therefore, H7 is partly confirmed by the analysis. The transitional generation engages the most in DNP (m=0.51, sd=0.77) and political consumerism (m=0.47, sd=0.73). War generation citizens exhibit the highest level of civic engagement compared to other generations (m=0.47, sd=0.88), and this level of engagement declines with each younger generation.
The B-F test found significant differences between generations in NML (F (4. 732)=2.61, p<.05), thus confirming hypothesis H5. The Games Howell post-hoc test showed that the post-war generation scores significantly lower on the NML scale than the post-transitional generation (p<.05). News media literacy linearly rises with younger generations and the youngest, post-transitional generation has the highest NML score (m=55.65, sd=5.83). There are no statistically significant differences in pluralist media orientation across generations, nor in the perception of misinformation online.
Table 3 demonstrates the results of the hierarchical regression analysis predicting institutional participation. The model explains 12,7% of the variance. Institutional participation is predicted by left–right placement (β = .11*), meaning that citizens who lean more towards the political right are more likely to be engaged in institutional politics. As shown in previous research, television use has a negative effect on participation (β = -.169*), while reading press media is a positive predictor (β = .143**). Neither participation in media literacy programmes nor the score on the news media literacy scale are significant predictors of institutional participation (H1 and H3 are rejected). Generational belonging also does not significantly predict institutional participation. However, citizens who perceive themselves as being less efficacious in recognising misinformation from online news are less prone to participate in institutional politics (β = - .167***).
*p<.05, **p<.01, ***p<.001
In the second regression analysis (Table 4), civic engagement is explained (12% of the variance is explained in the final model). Again, neither generational belonging nor the score in the news media literacy scale does not explain this type of participation. Civic engagement is positively predicted by reading press media (β = .18***) and participation in media literacy education programmes (β = .206***). Interestingly, the frequency of internet use is negatively correlated with civic engagement (β = -.124**). Again, the perception of having less self-efficacy in recognising misinformation is a negative predictor of this type of participation (β = -.101**).
*p<.05, **p<.01, ***p<.001
The third regression analysis (Table 5) predicts DNP (7,7% of the variance is explained). Here, gender is a predictor of this type of participation (β = -.148**), and men are more likely to participate in the digital sphere. DNP is also predicted by left–right placement (β = .095*), indicating that citizens aligning more closely with the political right tend to be more active in this way. Belonging to the socialist crisis generation (with the postwar generation as the reference category) is a significant predictor only in the third model (β = .147*). Finally, participation in media literacy programmes is positively correlated with DNP (β = .094*).
*p<.05, **p<.01, ***p<.001
In the final regression model, political consumerism is explained (7,7% of the variance). Here, we can see that generational belonging significantly predicts this kind of participation. More specifically, all generations younger than the postwar generation, except the youngest, post-transitional generation, engage more in political consumerism (socialist crisis generation, β = .254**; war generation, β = .226*; and transitional generation, β = .209*). Political consumerism is also negatively correlated with income (β = -.11*). This kind of participation is not predicted either by participation in media literacy programmes or by news media literacy. However, it is explained by pluralist media orientation (β = .099*).
*p<.05, **p<.01, ***p<.001
Discussion and Conclusions
The most significant result of this study is that news media literacy has no significant correlation with either type of political participation. As Vraga et al. (2021) claim, NML is not an end to itself, but its value is in its application which explains different related behaviours. There are several explanations for this result. As some of the earlier studies found, NML correlated with political efficacy, which usually correlates with higher political participation. However, the results of some studies also point to a correlation between NML and distrust towards the media and the political system (Ashley et al., 2017;Mihailidis, 2009;Tully and Vraga, 2018). Unfortunately, this study did not measure media trust and trust in institutions, but NML might be correlated with mistrust, which could explain the absence of correlation with political participation. Second, the NML scale might not work well enough in the sample of this study. Although the scale was tested across different samples in Ashley, Maksl and Craft (2013), and although the descriptives and the factor solution in this study resembled the original study, the scale might function differently in a different context. In addition, respondents scored highly on almost all the items on the scale, and there was no difference in NML between respondents who attended media literacy programmes and those who did not. Perhaps a scale that would address higher variability or one that addresses not only attitudes but also knowledge, as demonstrated by Ashley et al. (2017), could have explained the relationship with more nuance.
However, attending media literacy programmes was shown to be an important predictor of civic engagement and DNP. Even though NML or “civic media literacy” (Mihailidis and Thevenin, 2013) are not systematically taught in the Croatian education system, the results still point to the importance of such programmes for engaged citizenship. As many media literacy programmes in Croatia are provided sporadically by civil society initiatives (Kanižaj, 2016), respondents who participated in such programmes were already engaged to begin with. If this is true, this relationship could also be explained by a virtuous circle –citizens who are already engaged are perhaps more likely to have participated in media literacy programmes, which then supported their future engagement.
Levels of NML significantly differ across generations. It rises linearly with younger generations, and the youngest, post-transitional generation attains the highest NML score. Consistent with the findings of Vozab (2019b), the oldest generation possesses the lowest NML score. This points to a generational gap in news media literacy in Croatia, which should be addressed in future research.
One of the goals of this study was also to explain generational differences in distinct types of participation. Generations differ significantly in DNP and political consumerism. The results point to a rise in unconventional types of participation in each younger generation, which goes in line with some of the research in “old democracies” (e.g.,Marien, Hooghe and Quintelier, 2010). Unconventional forms of participation correlate with postmaterialist values and norms of engaged or self-actualising citizens (Bennett, Wells and Rank, 2009;Dalton, 2008). It is difficult to disentangle generational effects from life-cycle effects. However, if there is a linear trend across generations, it probably demonstrates the life-cycle effect (Andersen et al., 2021). These results point to the life-cycle effect, as DNP and political consumerism exhibit a linear rise among younger generations. According to Norris (2003), younger citizens are more drawn to unconventional participation because it is more flexible and addresses issues relevant to their experiences. However, the trend is shifting with the youngest, post-transitional generation, which is less engaged in unconventional types of participation than the transitional generation. It is difficult to conclude whether this is a generational trend or a life-cycle effect. As this generation is the youngest, perhaps this is just a sign of a lower political interest, which usually rises with age (Norris, 2003). On the other hand, if the trend is generational, it would suggest that this level of political participation might be more long-lasting.
Contrary to the presumption of the life-cycle perspective, in this study, older generations of citizens do not engage more in institutional participation. Norris (2003) argued that older age cohorts become more interested in politics and affiliate themselves more easily with political ideology and political parties, making them more active in institutional politics. The analysis found no significant differences among generations in institutional politics. Moreover, descriptive statistics show that younger generations are slightly more engaged than older citizens. Hooghe and Quintelier (2014) found in their analysis that older age cohorts in post-socialist countries participate less, likely due to socialisation in the one-party system. The results of this study also support this finding, as the two socialist generations engage the least in institutional participation.
This study has some limitations that should be mentioned. As already mentioned, this was the first t the NML scale was tested in this context and further expansion could provide a more nuanced understanding of its relationship with various behaviors. This is also a recommendation for future research. Future research could also consider interaction effects between news literacy and other variables, for example, media use. Although the quota sample of the study was designed to be representative of the adult population of Croatian citizens, some categories of respondents might be over or under-represented. Political participation was measured using a binary scale (“yes” and “no”), because of which some nuance might also be lost. Moreover, the question asked respondents whether they participated in a certain activity in the past year, which might not be fully representative of their behaviour. The survey was conducted during the COVID-19 pandemic, which influenced many behaviours, including news consumption and political participation. For this reason, the results of this study should be taken with caution, especially concerning the data from older respondents who were more vulnerable and might have been less inclined to engage in political activities.
Based on this study, practical recommendations for the development of media literacy in Croatia could be developed. First, as media literacy programmes are shown to be important for civic engagement and digitally networked participation, it would be important to strengthen the civic or NML dnsion in formal education, e.g., as advised by Mihailidis and Thevenin (2013). Second, although it is difficult to explain why NML is not correlated with any type of participation, it could be concerning that citizens with higher news media literacy are not more motivated to participate in politics. Therefore, NML programmes should acknowledge the possible correlation of NML with lower trust in the media and the political system and encourage participation. Third, the analysis has demonstrated a generational gap in NML. Media literacy is taught in formal education and many programmes target young citizens. Future media literacy policies and initiatives should address the generational gaps and offer educational programmes for older generations as well.
The results of this study are not unequivocally in line with the normative understanding of the relationship that media literacy should have with civic engagement. Although media literacy education programmes seem to have a positive impact on some forms of political participation, it remains unclear whether news media literacy fosters “healthy scepticism" or cynicism towards news. Considering the many risks that media literacy ought to alleviate in digital media environments (e.g., its role in policies combating disinformation,European Commission, 2018), there is a pressing need for a more nuanced approach to understanding the role of media literacy in a democracy.
FUNDING
This research was part of the thematic network "Media education is important (MOV)", (UP.04.2.1.06.0047). Total value of the project: HRK 2,975,852.47 (85% from the European Social Fund, and 15% from the Croatian state budget).
ETHICAL APPROVAL
Ethical approval was issued by the Institute for Development and International Relations in Zagreb (IRMO) on 26 January 2022.
DATA ACCESS AND TRANSPARENCY
Anonymised data and analytical material are available from the author upon request.
Footnotes:
[1] Some elements of the conceptualisation of political participation and age-related differences in participation in this article were also discussed in Vozab (2019a).
[2] Market research agency Valicon collected the data. The online sample consisted of respondents who were members of the agency's panel. For the face-to-face surveying, the sample was constructed using a snowball method while considering predefined socio-demographic quotas. This subsample was created to represent respondents who do not use the internet.
[3] Items were dummy coded as “yes” (1) and “no” (0) before summing them. For the sake of higher internal consistency, some variables from the extracted factors were omitted in the construction of composite variables. Therefore, civic engagement consists of three variables (Volunteered in a social or humanitarian organisation; Volunteered for a community project; Worked or volunteered for a civic action group). Institutional participation consists of two variables (Worked or volunteered for a party or candidate, Attended a meeting of a political party or other political organisation). Digitally networked participation consists of two variables (Posted or shared links on social media to political stories or articles for others to read; Commented on social media on political or social issues). Political consumerism consists of two variables (Boycotted certain products for political or ethical reasons; Deliberately bought certain products for political or ethical reasons).
[4] See the full list of items in Ashley, Maksl and Craft (2013).
[5] As the assumption of normality of distribution was violated for all the variables in the ANOVA analysis, the Brown-Forsythe Test was used.
[6] NML and pluralist media orientation are correlated at a statistically significant level r = .21, p < .001. There is no significant correlation between NML and perception of misinformation online. A T-test was performed, revealing no statistically significant differences in NML levels between respondents who attended education programmes in media literacy and those who did not.