Research article / 2025, Vol. 16, No. 1, pages 181-199
181
181
Consumo mediático y orientación política.
Un estudio sobre las Elecciones en España 2023
Authors:
Dimitria Jivkova-Semova Ángel Luis Rubio-Moraga Andrea Donofrio
Corresponding author:
Ángel Luis Rubio-Moraga alrubio@ucm.es
How to cite this article: Jivkova-Semova, D., Rubio- Moraga, A.L. & Donofrio, A. (2023). Media consumption and political orientation. A study on Elections in Spain 2023. Maskana, 16(2), 181–199. https://doi.org/10.18537/ mskn.16.01.12
doi: 10.18537/mskn.16.01.12
© Author(s) 2025. Attribution-NonCommercial- ShareAlike 4.0 International (CC BY-NC-SA 4.0)
Media consumption and political orientation. A study on Elections in Spain 2023
Consumo mediático y orientación política.
Un estudio sobre las Elecciones en España 2023
This study examines the bidirectional relationship between media consumption and ideological orientation in the context of Spain’s 2023 General Elections. Using a survey with 1,014 valid responses, the research explores how political preferences influence media selection and, conversely, how media consumption shapes ideological perspectives. It also assesses voters’ willingness to diversify their information sources, considering demographic variables such as age and gender.
The findings indicate that while most respondents prefer media aligned with their ideology, younger audiences tend to consult a wider range of sources. Only 17.7% reported having active subscriptions to paid press, with a preference for outlets perceived as neutral. Additionally, 5.9% actively sought media with opposing ideologies. The study concludes that confirmation bias remains dominant, although generational and gender differences influence media consumption patterns.
Keywords: media consumption, ideological positioning, elections, polarization, journalism.
El estudio analiza la relación bidireccional entre el consumo de medios y la orientación ideológica en el contexto de las Elecciones Generales de España de 2023. Mediante una encuesta con
1.014 respuestas válidas, se investigó cómo las preferencias políticas influyen en la selección de medios y viceversa. Se evaluó también la predisposición de los votantes a diversificar sus fuentes de información, considerando variables como edad y género.
Los resultados revelaron que, aunque la mayoría prefiere medios alineados con su ideología, los jóvenes muestran una mayor tendencia a consultar fuentes diversas. Solo el 17,7% tiene suscripciones a prensa de pago, con una inclinación hacia medios considerados neutrales. Además, un 5,9% buscó activamente medios con ideologías contrarias.
La investigación concluye que el sesgo de confirmación sigue predominando, pero existen diferencias generacionales y de género en el consumo mediático.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Introduction
The study of media consumption and its influence on ideological positioning has become an essential field of research in the social sciences and communication, being especially relevant in the context of elections, where voting decisions can be significantly influenced by exposure to different types of media. It is precisely for this reason that the present research focuses on the Spanish General Elections of 2023, an event of maximum relevance and crucial to understanding the contemporary dynamics between media, ideology, and electoral behavior.
First, the rationale for the topic lies in the need to understand how the media contributes to political polarization, defined as the process by which public opinions become more extreme and divided, which can have adverse effects on social cohesion and democratic governance. Numerous research studies have demonstrated that the media, particularly those with strong ideological biases, can reinforce individuals' existing beliefs, thereby limiting their exposure to opposing perspectives and exacerbating polarization (Stroud, 2010; Sunstein, 2001; Iyengar & Hahn, 2009; Gentzkow & Shapiro, 2011).
Stroud (2010) analyzed how individuals tend to choose sources of information that align with their own political beliefs, a phenomenon known as partisan selective exposure, which contributes to political polarization by systematically reducing their exposure to information that might challenge their views. For his part, Sunstein (2001) introduced the concept of "echo chambers" in his work to describe how individuals surround themselves with information that reinforces their existing beliefs, which can lead to the radicalization of their opinions and further social polarization. Iyengar and Hahn (2009) examined how preference for ideologically aligned media affected the perception of political reality among voters. They noted that individuals not only prefer media that reinforce their beliefs but also distrust those that present contrary perspectives, thus intensifying political polarization. Finally,
Gentzkow and Shapiro (2011) addressed ideological segregation in both online and traditional media, finding that segmentation of the information market based on consumer ideology is a key factor contributing to polarization.
In line with the latter research, it is worth noting that in an increasingly fragmented media environment characterized by the proliferation of digital platforms and social networks, citizens have an unprecedented ability to select the sources of information they consume. This capacity for personalization enables individuals to create "echo chambers," as referred to by Sunstein (2001), where they only encounter information that confirms their pre-existing beliefs, thereby generating a distorted view of political reality. The study of how this dynamic manifests itself during elections is crucial to understanding its impact on the democratic process. It is one of the main reasons that motivated the present study.
Current media consumption is deeply conditioned by the action of algorithms that select and hierarchize information in social networks, transforming these platforms into editorial actors without precise regulation (Cetina & Martínez Sierra, 2019). This algorithmic logic directly influences the way citizens access political information, favoring ideological bubbles and reducing informational diversity (López-López et al., 2020). Social media have also redefined the communication strategies of parties, promoting a more direct but also more polarized relationship with the electorate (Sierra & Rodríguez- Virgili, 2020). In this environment, large digital platforms act as invisible filters that determine which political discourses achieve visibility (Del Fresno & Daly, 2024).
In turn, political parties integrate automation and micro-segmentation processes that amplify their messages through algorithmic techniques (Campos-Domínguez & García-Orosa, 2018). This set of dynamics presents significant ethical and democratic challenges regarding
transparency, pluralism, and citizen participation in the digital public sphere.
Another key reason for conducting this research is the evolution of media consumption patterns among different demographic groups. Age and gender, in particular, are factors that can significantly influence media preferences and, thus, the formation of political opinions. Young people, for example, tend to consume a wider variety of information sources, including digital platforms and social networks, compared to older generations, who may prefer more traditional media such as television and print newspapers (Mitchell et al., 2016). Understanding these differences is crucial for developing effective political communication strategies that can effectively reach all segments of the population.
On the other hand, the specific context of the 2023 Spanish General Elections provides an ideal setting for this study for several reasons. First, these elections occurred during a period of high political volatility and significant shifts in the media landscape. The emergence of new political parties and increased distrust of traditional media have created an environment in which communication strategies and media consumption play a crucial role in shaping public opinion and voting decisions. Secondly, political polarization in Spain has increased in recent years, with an electorate increasingly divided between opposing ideological options (Camino- Rodríguez, 2021). This phenomenon not only affects social cohesion but also government effectiveness, as it hinders consensus building and the implementation of public policies. Between 2015 and 2023, Spain held four general elections, marked by notable fluctuations in the results and the emergence of new parties, which have significantly transformed the party system traditionally characterized by bipartisanship.
The emergence of these new formations in less than a decade reflects not only the growing ideological fragmentation of the electorate but also its sensitivity to increasingly polarizing and personalized media discourses. Their presence in the political landscape has been interpreted as a response to the crisis of representation, citizen disaffection (Bosch & Durán, 2017),
and the demand for more defined ideological alternatives. This plurality of actors has given rise to more polarized and media-intensive electoral campaigns, in which messages are tailored to specific niches of the electorate, and the media emerge as key channels for consolidating new political identities. In this new context of partisan reconfiguration, the media not only inform the public about the new options but also actively contribute to their legitimization or stigmatization, influencing the way the electorate perceives and evaluates them. Therefore, we believe that understanding how the media contributes to this polarization is fundamental to designing interventions that promote a more inclusive and constructive debate.
The state of the art reflects a rich tradition of research on the interaction between media and politics. From early work in the field of mass communication to more recent research on social networks, scholars have attempted to unravel how the media influences political attitudes and voting behavior. One of the most influential studies in this field is Lazarsfeld, Berelson, and Gaudet's The People's Choice (1948), a pioneering work that introduced the idea that the media have a limited impact on changing the political opinions of voters, who tend to select only information that confirms their pre-existing beliefs. This work contradicted the widespread perception that the media could significantly shape public opinions in a unidirectional manner. Wyatt (1998) critically reviewed the enduring influence of Lazarsfeld's work in his article "After 50 Years Political Communication Scholars Still Argue with Lazarsfeld," concluding that although technologies and contexts have changed, the fundamental ideas about the limited effects and influence of opinion leaders remain relevant. Wyatt also highlighted the need to consider the exogenous conditions and campaign events, as referred to by Holbrook (1996), in assessing the relative impact of election debates and other campaign events. Decades later, Lacy and Stamm (2015) reviewed and expanded on the concepts raised in Lazarsfeld's work and, like Wyatt, were able to corroborate the relevance of the original approaches in the modern context, further highlighting the persistence of the limited effects and influence of opinion leaders in the transmission of political information.
In the age of social networks, O'Regan's (2021) work offers a critical update on Katz and Lazarsfeld's personal influence theory. O'Regan argues that the dynamics of interpersonal communication identified by Katz and Lazarsfeld remain valid but are now manifested through social network interactions and the role of influencers. According to O'Regan, communication continues to occur in multiple stages, with opinion leaders and personal networks playing a crucial role in the dissemination and acceptance of political information. The concept of media polarization has also been widely studied, with a notable contribution by Stroud (2010), who explored how individuals tend to selectively expose themselves to information that reinforces their political ideologies, thereby intensifying political divisions by reducing exposure to opposing perspectives. For his part, as discussed above, Sunstein (2001) delved into the phenomenon of "echo chambers" created by the personalization of online information. Sunstein argues that the ability of users to select and filter content on the Internet enables the creation of homogeneous information environments, where individuals interact only with information that reinforces their pre-existing views. This phenomenon, known as "confirmation bias," has significant implications for political polarization and social cohesion.
Finally, the theoretical framework of agenda- setting, introduced by McCombs and Shaw (1972), is also relevant to this study. According to this theory, the media not only reflects reality but also influences public perception of the importance of issues. The authors demonstrated a strong correlation between the issues the media consider important and the issues the public perceives as priorities, suggesting that the media play an active role in shaping the public agenda. Additionally, the work of Zaller (1992) is noteworthy, as it offers a comprehensive understanding of how public opinions are formed and shaped. Zaller introduced the "R-A-S" (Receive-Accept-Sample) model, which describes how people receive, accept, and sample political information, highlighting the influence of ideological predisposition and exposure to media messages.
Nowadays, elections "are perhaps the closest thing to control of government by the people that can be achieved in modern industrialized mass society" (Milbrath, 1972, p. 154). During election periods, the role played by the media assumes special relevance as the flow of information intensifies and strategies of political persuasion are redefined. The media have become especially influential, both because of the aforementioned agenda-setting, which prioritizes specific issues in the public sphere, and because they frame the political debate in ideological terms, thereby configuring frameworks of interpretation that favor or penalize specific political proposals and actors. Media coverage during campaigns can alter the perception of candidates and key issues, affecting both the voting decision and the level of citizen participation. In this sense, contextualizing media consumption within the electoral cycle— in the weeks prior to the Spanish General Elections —allows us not only to understand the intensity of information use but also how electoral behavior is articulated through processes of selective exposure, framing, and polarization.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Subscription to paid media—whether digital or print—has become a significant variable for analyzing the relationship between citizens and the media, given that it implies an active commitment to specific editorial lines, levels of news quality, and journalistic sustainability models. Recent studies suggest that individuals who pay for access to news content tend to exhibit higher levels of political interest, a more critical attitude toward free information, and greater loyalty to media they perceive as credible or ideologically aligned. In this sense, "individuals change their source of information when they detect that the editorial line of the media has been modified," which revalidates "that political predispositions intervene in the loyalty of audience niches" (Humanes, 2014, p. 795).
Therefore, the inclusion of the subscription variable enables the analysis of not only the frequency of media exposure but also the perceived quality of the content and its relationship with the user's ideology, which justifies its incorporation as an analysis variable.
Several studies have highlighted that electoral periods are moments of intensified information, characterized by increased frequency of media consumption, which also becomes more selective and ideologically charged, justifying a more detailed analysis of media consumption under electoral pressure. The joint analysis of variables such as media subscription, content consumption, and ideological orientation aims to provide a comprehensive and nuanced view of the dynamics of information consumption in contemporary electoral processes.
As far as political ideology is concerned, it is one of the most studied concepts in political science. In a changing and polarized sociopolitical scenario such as the current one (McCoy et al., 2018), it is gaining increasing analytical relevance. In this study, political ideology is understood in terms of ideological self-placement on a left-right scale, a criterion widely used in research on polarization, electoral behavior, and selective exposure. Although the binary scheme of ideologies should be qualified due to its simplifying character in an increasingly fragmented and complex political reality, the scale remains a valid tool for capturing general ideological orientations and analyzing how they influence media consumption. In particular, its use allows us to examine the relationship between ideological affinity and media preferences, a key link to understanding the construction of information niches and loyalty to related media.
Based on an analysis of the state of the art, the following research objectives and hypotheses are proposed to comprehensively address the relationship between media consumption and ideological positioning during the 2023 Spanish General Elections. Thus, the primary objective is to examine the impact of media consumption on the ideological orientation of voters during the electoral process. This objective aims to comprehensively understand how media preferences influence ideological positions and how this relationship manifests in a specific electoral context.
The purpose of this study is to identify and describe trends in paid media subscriptions, as
well as general patterns of media consumption during the pre-election period. It also seeks to explore how media preferences relate to voters' political ideologies and to examine demographic differences in these patterns, paying special attention to variables such as age and gender. Another objective is to evaluate the impact of exposure to ideologically biased media on political polarization. Ultimately, the goal is to provide recommendations that foster more balanced and critical media consumption among voters.
The research is based on three main hypotheses:
H1. The first hypothesis posits a significant relationship between media consumption and the ideological positioning of voters in Spain. This hypothesis is based on previous studies that have shown how media preferences reflect and reinforce individuals' political ideologies. Research, such as Lazarsfeld et al. (1948) and Stroud (2010), has pointed to the tendency of voters to be exposed to media that confirm their pre-existing beliefs, suggesting a significant connection between media consumption and ideological stances.
H2. The second hypothesis suggests that media consumption patterns vary significantly by demographic factors such as age and gender. This hypothesis stems from observations in recent studies that reveal differences in media preferences among various demographic groups. For example, Mitchell et al. (2016) found that young people tend to diversify their information sources more compared to older generations, while Stroud (2010) identified differences in selective exposure based on gender.
H3. Finally, the third hypothesis proposes that exposure to media with strong ideological biases contributes to political polarization among voters. This hypothesis is supported by confirmation bias theory and Sunstein's (2001) research on "echo chambers," which suggests that voters who primarily consume media aligned with their ideologies tend to intensify their political beliefs, leading to greater polarization.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Materials and methods
The methodology of this study focuses on conducting a representative survey covering various demographic parameters, with special emphasis on gender and age. This survey was conducted in June and July 2023, coinciding with the weeks preceding the Spanish General Elections. The study design included 1014 valid surveys, distributed among six age groups: 18-24, 25-34, 35-44, 45-54, 55-64, and 65+, distributed according to the current percentage distribution:
- 18 to 24 years old: 10.3%.
- 25 to 34 years old: 13.5%.
From 35 to 44 years old: 19.2%.
From 45 to 54 years old: 18.9%.
From 55 to 64 years old: 15.3%.
Older than 65 years old: 22.8%.
To ensure the representativeness of the sample, a structured questionnaire was used, addressing key issues related to access to information, subscriptions to paid platforms, and the influence of political ideologies on media selection. The questionnaire was divided into two main blocks. The first block focused on general aspects of media consumption, including paid media subscriptions and ideological influence on media choice. The second block focused on the specific context of the July 2023 election campaign, using this juncture as a "laboratory" to examine the information dynamics and test the validity of the research hypotheses.
The data obtained were grouped and analyzed using absolute and relative frequencies, as well as statistical contrast methods, to evaluate the significance of the results. Two main statistical tests were employed: the Student's t-test and the chi-square test (χ²), both with a 95% confidence level. The Student's t-test is a fundamental statistical tool in this study. Its primary function is to compare the means of one or two samples with known population means or to directly compare the means between two samples (Härdel & Simar, 2019). The choice of this test is justified by its effectiveness in situations where sample sizes
are relatively small, and the variance between groups is not known a priori, which are common conditions in research across various disciplines. The application of the Student's t-test in this study followed a rigorous approach, starting with the verification of the assumptions necessary for its implementation. These assumptions include the normality of the data distribution and the homogeneity of variances between the compared samples.
The results of the Student's t-test provide crucial evidence on the presence or absence of statistically significant differences in the analyzed data. A p-value less than 0.05 (for a 95% confidence level) indicates that the differences between the means are statistically significant, suggesting that these differences are unlikely to be due to chance. Conversely, a p-value greater than 0.05 suggests that there is insufficient evidence to claim a significant difference between the means compared. Careful interpretation of these results is essential to draw valid conclusions from this study. The statistical significance identified through the Student's t-test enables the detection of relevant patterns and trends within the data, significantly contributing to the understanding of the underlying dynamics in the research area.
The chi-square test (χ²), on the other hand, is a nonparametric statistical test widely used to investigate whether distributions of categorical variables differ from theoretical or expected distributions. In the context of this research, the chi-square test has been employed to evaluate hypotheses regarding the relationship between categorical variables, providing a rigorous method for determining whether observed differences between data segments are statistically significant. The chi-square test procedure begins with the formulation of a contingency table that summarizes the observed frequencies of the events or characteristics studied, distributed in various categories.
The calculated χ² value is then compared with a critical value of the chi-square distribution, which depends on the desired confidence level (commonly 95%) and the degrees of freedom of the test (typically the number of categories minus one). If the χ² value exceeds the critical value, the null hypothesis is rejected, indicating that there are statistically significant differences between the observed and expected frequencies. This interpretation is crucial for determining the validity of the hypotheses put forward and for better understanding the relationships between the variables studied.
The questionnaire used in this study consisted of two main blocks. The first block was dedicated to the issue of access to information and subscriptions to paid platforms and newspaper paywalls, as well as the influence of political ideologies on media selection. This block also covered local press, radio, and television consumption, allowing for a comprehensive assessment of respondents' media behavior. The second block used the July 2023 election campaign in Spain as a contextual framework to test all questions related to news flows in a particular situation. This approach enables us to examine how informative dynamics impact electoral behavior, utilizing the campaign as a natural laboratory to test the validity of hypotheses and research questions.
In summary, the methodology of this study is based on a representative survey design that encompasses key demographic parameters and uses robust statistical methods to analyze the data. The combination of the Student's t-test and the chi-square test allows for a detailed and rigorous assessment of the significant differences and relationships between the variables studied, providing an in-depth understanding of the dynamics between media consumption and ideological positioning in the context of the 2023 Spanish General Elections. This methodological approach ensures that study findings are valid and reliable, making a significant contribution to academic knowledge on the influence of the media in contemporary politics.
The methodology of this study focuses on conducting a representative survey covering various demographic parameters, with special
emphasis on gender and age. This survey was conducted in June and July 2023, coinciding with the weeks preceding the Spanish General Elections. The study design included 1014 valid surveys, distributed among six age groups: 18-24, 25-34, 35-44, 45-54, 55-64, and 65+, distributed according to the current percentage distribution:
- 18 to 24 years old: 10.3%.
- 25 to 34 years old: 13.5%.
From 35 to 44 years old: 19.2%.
From 45 to 54 years old: 18.9%.
From 55 to 64 years old: 15.3%.
Older than 65 years old: 22.8%.
To ensure the representativeness of the sample, a structured questionnaire was used, addressing key issues related to access to information, subscriptions to paid platforms, and the influence of political ideologies on media selection. The questionnaire was divided into two main blocks. The first block focused on general aspects of media consumption, including paid media subscriptions and ideological influence on media choice. The second block focused on the specific context of the July 2023 election campaign, using this juncture as a "laboratory" to examine the information dynamics and test the validity of the research hypotheses.
The data obtained were grouped and analyzed using absolute and relative frequencies, as well as statistical contrast methods, to evaluate the significance of the results. Two main statistical tests were employed: the Student's t-test and the chi-square test (χ²), both with a 95% confidence level. The Student's t-test is a fundamental statistical tool in this study. Its primary function is to compare the means of one or two samples with known population means or to directly compare the means between two samples (Härdel & Simar, 2019). The choice of this test is justified by its effectiveness in situations where sample sizes are relatively small, and the variance between groups is not known a priori, which are common conditions in research across various disciplines. The application of the Student's t-test in this study followed a rigorous approach, starting with the verification of the assumptions necessary for its implementation. These assumptions include the normality of the data distribution and the
homogeneity of variances between the compared samples.
The results of the Student's t-test provide crucial evidence on the presence or absence of statistically significant differences in the analyzed data. A p-value less than 0.05 (for a 95% confidence level) indicates that the differences between the means are statistically significant, suggesting that these differences are unlikely to be due to chance. Conversely, a p-value greater than 0.05 suggests that there is insufficient evidence to claim a significant difference between the means compared. Careful interpretation of these results is essential to draw valid conclusions from this study. The statistical significance identified through the Student's t-test enables the detection of relevant patterns and trends within the data, significantly contributing to the understanding of the underlying dynamics in the research area.
The chi-square test (χ²), on the other hand, is a nonparametric statistical test widely used to investigate whether distributions of categorical variables differ from theoretical or expected distributions. In the context of this research, the chi-square test has been employed to evaluate hypotheses regarding the relationship between categorical variables, providing a rigorous method for determining whether observed differences between data segments are statistically significant. The chi-square test procedure begins with the formulation of a contingency table that summarizes the observed frequencies of the events or characteristics studied, distributed in various categories.
The calculated χ² value is then compared with a critical value of the chi-square distribution, which depends on the desired confidence level (commonly 95%) and the degrees of freedom
of the test (typically the number of categories minus one). If the χ² value exceeds the critical value, the null hypothesis is rejected, indicating that there are statistically significant differences between the observed and expected frequencies. This interpretation is crucial for determining the validity of the hypotheses put forward and for better understanding the relationships between the variables studied.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
The questionnaire used in this study consisted of two main blocks. The first block was dedicated to the issue of access to information and subscriptions to paid platforms and newspaper paywalls, as well as the influence of political ideologies on media selection. This block also covered local press, radio, and television consumption, allowing for a comprehensive assessment of respondents' media behavior. The second block used the July 2023 election campaign in Spain as a contextual framework to test all questions related to news flows in a particular situation. This approach enables us to examine how informative dynamics impact electoral behavior, utilizing the campaign as a natural laboratory to test the validity of hypotheses and research questions.
In summary, the methodology of this study is based on a representative survey design that encompasses key demographic parameters and uses robust statistical methods to analyze the data. The combination of the Student's t-test and the chi-square test allows for a detailed and rigorous assessment of the significant differences and relationships between the variables studied, providing an in-depth understanding of the dynamics between media consumption and ideological positioning in the context of the 2023 Spanish General Elections. This methodological approach ensures that study findings are valid and reliable, making a significant contribution to academic knowledge on the influence of the media in contemporary politics.
Results
The analysis of the data collected through the survey reveals several significant findings regarding the media consumption and ideological positioning of voters in the 2023 Spanish General Elections. In this section, we will break down these results, presenting both general trends and specific differences according to different demographic parameters, such as age and gender.
Of the total number of respondents, only 17.7% reported having an active subscription to paid newspapers. Figure 1 illustrates the breakdown
of these data by ideological positioning. It shows that nearly half of the respondents (48%) indicated, according to their assessment, that they do not have subscriptions to media with a clear ideological stance. Only 10% admitted to subscribing to publications aligned with their ideology. This finding suggests that a large majority of respondents prefer media that they perceive as neutral or do not identify with a specific ideology, which may indicate a search for objectivity or a distrust of clearly ideologized media.
Figure 1: Subscription to press and platforms from the perspective of ideological positioning.
Source: Own elaboration
When analyzing the responses from the perspective of age, we find notable differences. Thus, people between 18 and 24 years of age tend to subscribe more to media with ideologies contrary to their own, contrasting with those in the 55 to 64 age range, who show the least inclination to do so (Table 1). This pattern may reflect a greater openness and desire for exposure to diverse perspectives among young people, possibly influenced by their increased use of the Internet and social networks, where it is more common to encounter a variety of viewpoints. In contrast, older age groups may show a stronger preference for ideological consistency and confirmation of their pre-existing beliefs.
Another important aspect of the study is the total media consumption during the General Elections, which includes paid media as well
as print and free-to-air mass media (MCM). The results indicate that 59% of respondents admitted to having sought out MCM with opposing ideologies to contrast or expand their perspective. In comparison, 12.7% claimed to have consulted sources other than their political ideals specifically to reaffirm their own beliefs. This finding highlights the presence of a substantial group of voters who actively seek to balance their information intake. However, most people still prefer to consume media that aligns with their ideologies.
In terms of gender, there are apparent differences. Women tend to consume more MCMs that align with their ideology compared to men, and in turn, they are less likely to turn to ideologically opposed sources.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Table 1: Total media consumption during the last General Elections (pay wall + press and free-to-air MCM).
Source: Own elaboration based on the survey.
Significance Level: 95%. | Total | Man | Woman |
Total | 1014 | 498 | 516 |
I have only consumed information in media close to my ideology. | 22,2 | 18.7- | 25.6+ |
I have consumed, for the most part, media close to my ideology, but I have also sought some information in neutral media that I had not previously consumed. | 22,5 | 23,3 | 21,7 |
I have consumed, for the most part, media close to my ideology, but I have also sought some information in contrary media that I had not previously consumed. | 22,2 | 26.9+ | 17.6- |
I discovered new media and found them interesting, without thinking about their ideology. | 14,5 | 14,1 | 14,9 |
I have discovered new media, contrary to my ideology, and I thought it was interesting to contrast. | 5,9 | 5,6 | 6,2 |
I have searched for media contrary to my ideology, just to confirm my opinion. | 12,7 | 11,4 | 14,0 |
This pattern may reflect differences in how men and women process information and seek confirmation of their beliefs. Additionally, it may be influenced by various social expectations and gender roles that impact media exposure and consumption.
Contrary to what might be expected, significant differences by age are minimal, with one notable exception: voters aged 45-54 are more likely to consult sources they perceive as neutral. This trend stands out, especially in comparison to the 35-44 and 55-64 age groups, where interest in these types of sources is lower. This finding suggests that middle-aged individuals may be at a stage in their lives where they seek to balance diverse information and opinions, perhaps due to their work and family responsibilities that require a broader and more nuanced understanding of political issues.
A control question included in the survey was,
"Have you followed more media and newspapers
Discussion
that you usually do not follow during the election campaign?" The result confirms the findings described above, as 77% of voters stated that the election campaign had not affected their usual MCM consumption. This suggests that, although elections generate an increase in news activity, most voters tend to maintain their established consumption patterns, which may reflect the depth of confirmation bias and resistance to change in media preferences.
The results of this survey provide a detailed insight into how Spanish voters consume media about their ideological orientation. Demographic differences in media preferences and the general tendency to seek information that confirms pre- existing beliefs underscore the importance of considering these factors when analyzing the media's impact on politics. These findings not only contribute to academic knowledge about the influence of media but also have practical implications for the formulation of more effective communication strategies and the promotion of more critical and balanced media consumption.
The results of this research confirm the existence of a significant relationship between media consumption and the ideological orientation of voters in the context of the 2023 General Elections in Spain. The data indicate that, although the majority of respondents prefer to
consume media aligned with their ideology, there is a tendency among young people to diversify their information sources. This finding suggests that new generations may be more willing to be exposed to different political perspectives, which contrasts with the trend observed in
previous studies, such as Valera-Ordaz et al. (2021), Stroud (2010), and Sunstein (2001), who described a reinforcement of confirmation bias and the formation of "echo chambers" in news consumption patterns. The difference observed in this research could be linked to the transformation of the media ecosystem, where digital platforms and social networks have diversified the information options available and have facilitated access to content of different ideological tendencies. However, this apparent informative openness should be interpreted with caution since exposure to diverse media does not necessarily guarantee greater plurality in the construction of political opinions. Studies, such as those by Mitchell et al. (2016), have pointed out that, although young people access a variety of information sources, their consumption may be conditioned by recommendation algorithms that reinforce their previous preferences, thereby limiting the impact of media diversity. In this sense, although the data suggest that young people have a greater predisposition to consult sources with different ideological approaches, it cannot be stated with certainty that this leads to a reduction in political polarization, as suggested by Waisbord (2020).
In terms of methodological comparison, this study aligns with previous research that has employed surveys as a primary tool to assess media consumption and its relationship with political ideology (Gentzkow & Shapiro, 2011; Endersby, 2011). However, a key difference is that this research has specifically explored the dynamics of information consumption during a specific election period, which provides a more contextualized view of how citizens access information at times of high political polarization. Unlike longitudinal studies that analyze long-term trends (Iyengar & Hahn, 2009), this research focuses on a specific event, which allows us to identify consumption patterns in a small but representative time frame. This approach has both strengths and limitations. On the one hand, it provides a detailed snapshot of how voters consume information in an electoral context, enabling us to analyze the immediate effects of media coverage and the influence of political discourse on the construction of public opinion. On the other hand, focusing on a specific
moment does not allow us to assess whether these consumption patterns remain stable over time or whether they are subject to variations depending on the political and media context.
Another relevant methodological difference in previous studies lies in the way generational, and gender differences in the selection of information sources have been addressed. While previous research has analyzed media consumption in general terms, this study has placed particular emphasis on how these demographic factors influence the diversification or restriction of voters' information diet. The identification of trends, such as the greater predisposition of young people to consume diverse sources and the lesser inclination of women to seek out media with opposing ideologies, provides a more nuanced understanding of the phenomenon, complementing the contributions of authors like Pereira-López et al. (2023). These differences suggest that political communication strategies and media literacy should consider not only the ideology of voters but also the sociodemographic characteristics that influence their information habits.
In addition, this research presents a detailed analysis of the role of paid versus free media, an aspect that has been less explicitly addressed in previous studies. The data obtained show that citizens who opt for subscriptions to paid media tend to prefer those considered neutral or without a clear ideological affiliation, which opens new questions about the impact of media funding on the perception of media credibility and the construction of public debate. This finding raises the need for further research on whether media with subscription models exert a different influence on the formation of public opinion compared to those that rely exclusively on advertising or open access.
Regarding the interpretation of the data, the results show that only 17.7% of respondents have active subscriptions to paid press, which raises questions about the sustainability of quality journalism in an information ecosystem where free and open-access sources predominate. This finding is consistent with previous research that has pointed to a decrease in the willingness to
pay for news in digital contexts (Gentzkow & Shapiro, 2011). However, the novelty of this research lies in the differentiation between free and paid media consumption according to voter ideology, which allows inferring that media considered neutral have a higher probability of attracting subscribers than those with a clearly defined political orientation. This data suggests that the perception of impartiality remains a significant factor in trust in the media, an aspect that has been widely debated in studies on journalistic credibility (Tsfati & Cappella, 2003; Masip et al., 2020).
The fact that most voters prefer to consume media without an explicit ideological orientation opens a debate about the concept of objectivity in contemporary journalism. While some authors have argued that the media should assume a clear editorial stance to avoid false equivalence between unequal positions (Bennett, 2016), the results of this study suggest that a significant part of the public seeks information in media they perceive as unbiased. However, this does not necessarily imply less polarization, as the perception of neutrality may be mediated by the recipient's own beliefs. Research, such as that of Gunther (1992), has shown that individuals tend to consider as "objective" those media that align with their ideological framework, which could explain why certain media with a defined editorial line continue to be perceived as neutral by some segments of the audience.
Another point to highlight is the influence of free access on the selection of information media. Access to content without economic restrictions could be affecting voters' willingness to pay for quality information, which represents a challenge to the sustainability of independent journalism. In this sense, studies such as those by Chyi and Yang (2009) have shown that, although citizens value quality information, they are rarely willing to pay for it if they have free access to news through social networks or digital platforms. This has important implications for the future of journalism, as subscription-based business models may not be viable if the public does not perceive significant added value in paid content. In turn, the prevalence of free media with advertising-based funding models may
incentivize the production of content designed to maximize virality rather than informative depth, which could contribute to the spread of biased or decontextualized news.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Ultimately, the low rate of paid media subscriptions reflects not only a preference for free media but also a trend toward information diversification inaccessible platforms. The apparent contradiction between the search for neutral media and the persistence of confirmation bias requires a deeper analysis of how voters construct their perception of impartiality in today's media environment. The increasing influence of algorithms in news distribution and content personalization may significantly shape this perception, suggesting the need for future research that combines quantitative and qualitative methods to assess the relationship between trust in media, willingness to pay for information, and consumption habits in the digital age.
A relevant aspect to consider is the low proportion of citizens who actively seek information in media with opposing ideologies (5.9%). This data reinforces the hypothesis that selective exposure is still predominant in most voters, which coincides with studies such as those of Lazarsfeld et al. (1948), Lacy and Stamm (2017), Valera-Ordaz (2023) and Doménech-Beltrán (2024), who pointed out that voters tend to expose themselves to information that reinforces their pre-existing beliefs. However, the existence of a small percentage of voters who deliberately consume media contrary to their ideology could indicate a possible evolution in information habits that deserves further exploration in future studies.
This finding is relevant in the context of the debate on media polarization and fragmentation of the information ecosystem. Most theoretical models have suggested that information consumption is organized around "echo chambers" (Sunstein, 2001), where citizens primarily interact with content aligned with their values and beliefs, thereby reinforcing their political identity and limiting exposure to divergent perspectives. However, the presence of a sector, albeit a minority, of the population that deliberately
seeks out media with a different ideological orientation, could suggest the existence of a still incipient disposition towards a more pluralistic media consumption. This behavior is especially relevant in the case of young voters, who, as the data in this study indicate, tend to diversify their information sources more than older generations.
The fact that a small percentage of voters actively consume opposing media can be explained from different perspectives. One possible interpretation is that these individuals seek to contrast information to reinforce their position rather than being open to changing their minds. Previous research has shown that exposure to contradictory information does not always lead to a change in attitude. However, in some cases, it may further strengthen pre-existing beliefs, a phenomenon known as the "backfire effect" (Nyhan & Reifler, 2010). Another possibility is that these mostly young voters face a more digitalized and less ideologically structured media environment, where access to diverse information is easier than in traditional formats.
Among the anomalous results is the fact that 77% of the respondents affirm that their media consumption did not vary during the electoral campaign. This finding contrasts with previous literature, suggesting that election periods tend to increase the search for political information (McCombs & Shaw, 1972). One possible explanation lies in the increasing personalization of media consumption, driven by algorithms that filter information based on a user's interests. Alternatively, this could suggest a growing disinterest in traditional media coverage of political campaigns, raising questions about the role of the media in electoral mobilization and shaping public opinion.
In terms of novel contributions, this research highlights the importance of generational differences in media selection and the impact of digitization on exposure to ideologically diverse information. The tendency of young people to seek out more varied sources could be interpreted as an indicator that future generations will be less influenced by confirmation bias and more open to informational diversity. This aspect merits further investigation in future studies.
Conclusions
In light of the research results, and based on the proposed objectives and hypotheses, as well as the obtained findings, several conclusions can be drawn that encompass both theoretical and practical aspects.
The first hypothesis, which posited a significant relationship between media consumption and voters' ideological positioning, is confirmed by the data obtained. The results show that a large majority of respondents prefer to consume media that they perceive as neutral, although a considerable percentage choose media aligned with their ideology. This reinforces the theory of confirmation bias, which suggests that individuals tend to seek information that reinforces their pre-
existing beliefs. This finding is consistent with previous studies, such as those by Lazarsfeld et al. (1948) and Stroud (2010), which have demonstrated how voters tend to select information that confirms their existing views. In this context, media consumption not only reflects existing ideologies but can also intensify them, contributing to political polarization. Voters who actively seek out media with opposing ideological biases represent a minority, indicating a general tendency toward consolidation of pre-existing beliefs rather than openness to new perspectives.
The second hypothesis proposed that media consumption patterns vary significantly according to demographic factors such as age and gender.
The results confirm this hypothesis, showing apparent differences in media preferences between different demographic groups. In terms of age, 18-24-year-olds tend to subscribe to media with ideologies contrary to their own more than any other age group. This trend may reflect a greater openness and desire for exposure to diverse perspectives among young people, influenced by their increased use of the Internet and social networks. In contrast, older age groups, especially those between 55 and 64, tend to show a lesser inclination to consume media with opposing ideologies, instead preferring media aligned with their own beliefs.
Gender analysis reveals that women tend to consume more media aligned with their ideology compared to men and are less likely to seek out ideologically opposing sources. This difference may be related to the distinct ways in which men and women process information and seek confirmation of their beliefs, as well as social expectations and gender roles that influence their media consumption.
The third hypothesis suggested that exposure to media with marked ideological biases contributes to political polarization among voters. The results obtained indicate that a significant 66.9% of voters prefer media with related ideologies, although a considerable portion also explores options they consider neutral. The general tendency towards consuming media that aligns with one's ideologies confirms the theory of confirmation bias and "echo chambers" described by Sunstein (2001). This practice can intensify existing political beliefs and contribute to further polarization, as individuals constantly receive information that reinforces their views and is rarely exposed to opposing perspectives.
These findings also enable us to explore the transformations of the information ecosystem about perceptions of media impartiality and the value assigned to paid media. One of the most noteworthy findings is that, despite only 17.7% of those surveyedsubscribinge to paid press, there is a clear preference for media perceived as neutral, suggesting that the subscription model
could remain viable if it guarantees quality and editorial independence. A key finding from the control question is that 77% of voters reported that the election campaign did not alter their usual media consumption. This suggests that, despite the intensified news activity during elections, voters tend to maintain their established consumption patterns. This finding reflects the depth of confirmation bias and resistance to change in media preferences, indicating that election campaigns have a limited impact on altering voters' information consumption habits.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
The findings of this study have several practical implications for political communication and the formulation of media strategies. First, the results suggest the need to promote a more balanced and critical media consumption among voters. Given that most voters tend to seek information that confirms their pre-existing beliefs, it is crucial to develop educational programs and awareness campaigns that encourage critical evaluation of information and exposure to diverse perspectives. Additionally, policymakers and communication professionals should consider demographic differences in media preferences when developing communication strategies. For example, campaigns targeting young people could benefit from a greater presence on digital platforms and social networks. In contrast, campaigns for older age groups should focus more on traditional media and ideological consistency in messaging.
Likewise, this study opens new lines of research focused on the longitudinal analysis of news behavior during different electoral processes, as well as evaluating the effectiveness of media literacy programs in counteracting confirmation bias. In addition, the results prompt us to consider the importance of cultivating a media-critical citizen who can recognize ideological biases in news content and value plurality as a crucial element in democratic societies. Ultimately, it is crucial to address political polarization by fostering dialogue and interaction among individuals with diverse perspectives. The media can play a key role in this effort by providing balanced and diverse coverage, avoiding the creation of "echo chambers" that only reinforce existing divisions.
Recommendations
The findings of this research highlight the influence of media consumption in the consolidation of ideological beliefs and its role in political polarization. Based on these results, several recommendations can be made both in the academic field and in the design of communication strategies and public policies. First, it would be advisable to encourage a more balanced media consumption through media literacy programs that promote exposure to diverse sources. Additionally, the development of tools that facilitate the comparison of news from different perspectives could help mitigate confirmation bias. Second, there is a need to investigate other factors that influence media selection, such as educational level, socioeconomic context, and the use of social networks, which could modify these dynamics. Likewise, given that this study focuses on a specific electoral moment, it would be beneficial to conduct longitudinal analyses that enable us to evaluate the evolution of news habits across different political and social contexts. From a methodological perspective, the
inclusion of qualitative approaches, such as in- depth interviews or focus groups, would facilitate a deeper understanding of the decision-making processes involved in selecting information sources. Likewise, it is worth noting that this work has certain limitations, including its quantitative approach and focus on a single electoral context, which invites future research to broaden the time frame and incorporate comparative designs that enhance the external validity of the results. Finally, it is essential to examine the impact of new digital formats, in particular social networks and recommendation algorithms, on the configuration of political orientation. To this end, experimental studies could assess whether exposure to diverse news in these environments has a real impact on reducing polarization. In conclusion, future research should consider mixed methodologies and comparative approaches to broaden the understanding of the link between media and ideology. From an applied perspective, it is essential to design educational and political strategies that encourage more critical and pluralistic media consumption.
Acknowledgments
The present research has been developed within the framework of the R&D Project PR27/21-017 TTAI, entitled "Traceability, Transparency, and Access to Information: Study and Analysis of the Dynamics and Trends in the Field," funded by the Community of Madrid within the R&D Projects Program of the UCM 2021 Plan. This project,
coordinated between Spain, Portugal, and the United Kingdom and directed by the researcher Graciela Padilla Castillo from the Complutense University of Madrid, has as its primary objective the study of the traceability of information at the European level in order to combat disinformation and ensure access to truthful and contrasted information for citizens.
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
Bibliographic references
Bennett, W. L. (2016). News: The Politics of Illusion (10th Edition). Editorial University Of Chicago Press.
Bosch, A., & Durán, I. M. (2017). How does economic crisis impel emerging parties on the road to elections? The case of the Spanish Podemos and Ciudadanos. Party Politics, 25(2), 257-267. https://doi.org/10.1177/1354068817710223.
Camino-Rodríguez, E. (2021). Ideological, electoral and affective polarization in Spain (2008-2019). [Master's Thesis, Universidad Complutense de Madrid]. https://www.ucm.es/ comunicacion-politica/file/tfm-eduardo-camino
Campos-Domínguez, E., & García-Orosa, B. (2018). Algorithmic communication in political parties: automation of message production and circulation. El Profesional de la Información, 27(4), 778-786. https://doi.org/10.3145/epi.2018. jul.06
Cetina Presuel, R., & Martínez Sierra, J. M. (2019). Algorithms and news: Social networks as news publishers and distributors. Journal of Communication, 18(1), 75-97. https:// revistadecomunicacion.com/article/view/1297.
Chyi, H. I., &Yang, M. J. (2009). Is Online News an Inferior Good? Examining the Economic Nature of Online News among Users. Journalism & Mass Communication Quarterly, 86(3), 594-612. https://doi.org/10.1177/107769900908600309
Del Fresno García, M., Daly, A. J. (2024). Limits to political communication from large Internet social platforms. A case study from "la casta" to "la trama". Revista Española de Investigaciones Sociológicas, 185, 117-136. https://reis.cis.es/ index.php/reis/article/view/799.
Doménech-Beltrán, J. (2024). Cross-sectional exposure to the Spanish public broadcasting system: influence of ideology, partisanship and interest in politics on the consumption of RTVE,
Communication & Society, 37(2), 19-33. https:// revistas.unav.edu/index.php/communication- and-society/article/download/45006/40105/
Endersby, J. (2011). Fair and Balanced? News Media Bias and Influence on Voters. SSRN Electronic Journal. 10. https://dx.doi. org/10.2139/ssrn.1920620
Gentzkow, M., & Shapiro, J. M. (2011). Ideological Segregation Online and Offline. Quarterly Journal of Economics, 126(4), 1799-
1839. https://doi.org/10.1093/qje/qjr044
Humanes, M. L. (2014). Selective exposure and audience partisanship in Spain: The consumption of political information during the 2008 and 2011 election campaigns. Palabra Clave, 17(3), 773-
802. http://dx.doi.org/10.5294/pacla.2014.17.3.9
Gunther, A. C. (1992). Biased Press or Biased Public? Attitudes Toward Media Coverage of Social Groups. Public Opinion Quarterly, Volume 56, Issue 2, SUMMER, 147-167, https:// doi.org/10.1086/269308
Härdel, W. K., and Simar, L. (2019). Applied Multivariate Statistical Analysis (5th ed.). Springer.
Holbrook, T. M. (1996). Do Campaigns Matter? SAGE Publications. https://doi. org/10.4135/9781452243825
Iyengar, S., & Hahn, K. S. (2009). Red Media, Blue Media: Evidence of Ideological Selectivity in Media Use. Journal of Communication, 59(1), 19-39. https://doi.org/10.1111/j.1460-
2466.2008.01402.x
Katz, E., and Lazarsfeld, P. F. (1964). Personal Influence, the Part Played by People in the Flow of Mass Communications. Free Press.
Lacy, S., & Stamm, M. (2017). Reassessing the People's Choice: Revisiting a Classic and
Excavating Lessons for Research About Media and Voting. In R. Wei, Advances in Foundational Mass Communication Theories. Routledge.
Lazarsfeld, P. F., Berelson, B., and Gaudet, H. (1948). The People's Choice. How the Voter Makes Up His Mind in a Presidential Campaign. Columbia University Press.
López-López, P. C., Jaráiz, E., & Crespo-Pereira,
V. (2020). The influence of social networks on media and political participation. Revista Ibérica de Sistemas e Tecnologias de Informação, E35, 76-87. https://www.researchgate.net/ publication/339090703.
Martínez-Vares, S. (2025). "What the politician says and what people understand". Interview in La Ventana Andalucía, Cadena SER. https:// cadenaser.com/andalucia/2025/02/18/martinez- vares-frente-a-la-antipolitica-es-importantisimo- que-los-politicos-hablen-de-las-cosas-que-le- importan-a-la-gente-radio-sevilla/
Masip, P., Suau, J., & Ruiz-Caballero, C. (2020). Media perceptions and disinformation: ideology and polarization in the Spanish media system. Profesional De La información,29(5). https:// doi.org/10.3145/epi.2020.sep.27
McCoy, J., Rahman, T., & Somer, M. (2018). Polarization and the global crisis of democracy: Common patterns, dynamics, and pernicious consequences for democratic policies. American Behavioral Scientist, 62(1), 16-42. https://doi. org/10.1177/0002764218759576
Milbrath, L. W. (1972). Political Participation.
Rand McNally & Com.
McCombs, M. E., and Shaw, D. L. (1972). The Agenda-Setting Function of Mass Media. The Public Opinion Quarterly, 36(2), 176-187. http:// www.jstor.org/stable/2747787
Mitchell, A., Gottfried, J., Barthel, M., & Shearer,
E. (2016). The Modern News Consumer: News attitudes and practices in the digital era. Pew Research Center. https://www.pewresearch. org/journalism/2016/07/07/the-modern-news- consumer/
Nyhan, B., & Reifler, J. (2010). When corrections fail: The persistence of political misperceptions. Political Behavior, 32(2), 303–330. https://doi. org/10.1007/s11109-010-9112-2
O'Regan, T. (2021). Re-Reading Personal Influence in an Age of Social Media. Journal of Communication, 71(3), 437-455. https://doi.org/10.3366/ edinburgh/9781474474467.003.0010
Pereira López, M., Vicente Iglesias, G., and García Maseda, M. (2023). Mujeres, consumo de medios y extrema derecha: Women, media consumption and the far right. Razón Y Palabra, 27(118), 90–104. https://doi.org/10.26807/ rp.v27i118.2082
Sierra, A., and Rodríguez-Virgili, J. (2020). Political communication and social media. Revista Mediterránea de Comunicación, 11(1), 185-198. https://www.mediterranea-comunicacion.org/ article/view/2020-11-1-comunicacion-politica- y-medios-sociales
Stroud, N. J. (2010). Polarization and Partisan Selective Exposure. Journal of Communication, 60(3), 556-576. https://doi.org/10.1111/j.1460-
2466.2010.01497.x
Sunstein, C. R. (2001). Echo Chambers: Bush
v. Gore, Impeachment, and Beyond. Princeton University Press.
Tsfati, Y., & Cappella, J. N. (2003). Do People Watch What They Do Not Trust? Exploring the Association between News Media Skepticism and Exposure. Communication Research, 30, 504-
529. https://doi.org/10.1177/0093650203253371
Valera Ordaz, L., Requena Mora, M., & Pop,
A. I. (2021). Ideology and selective exposure: the media consumption of Spaniards during the April 2019 campaign. In G. López García and E. Campos Domínguez (Eds.), Redes en campaña: análisis de la comunicación política en las elecciones generales de 2019 (pp. 157-173). Comunicación Social Ediciones y Publicaciones. https://doi.org/10.52495/c9.emcs.24.p93
Valera-Ordaz, L. (2023). Research on selective media exposure in Spain: a critical review of
Dimitria Jivkova-Semova, Ángel Luis Rubio-Moraga, Andrea Donofrio
its findings, application phases, and blind spots. Profesional De La Información, 32(5). https:// doi.org/10.3145/epi.2023.sep.07
Waisbord, S. (2020) Is it valid to attribute political polarization to digital communication? On bubbles, platforms and affective polarization. SAAP Magazine: Argentine Society of Political Analysis, 14(2), 249-279. https://doi. org/10.46468/rsaap.14.2.A1
Wyatt, R. O. (1998). After 50 Years, Political Communication Scholars Still Argue with Lazarsfeld. Journal of Communication, 48(2), 146-156. https://doi. org/10.1111/j.1460-2466.1998.tb02752.x
Zaller, J. R. (1992). The Nature and Origins of Mass Opinion. Cambridge University Press. https://doi.org/10.1017/CBO9780511818691