Sociodemographic profiles and predictors of news consumers
Composición y predictores sociodemográficos de los consumidores de noticias

Juan Pablo Artero1
Víctor Orive1
Pilar Latorre1

1University of Zaragoza. Spain.

Introduction: This research article examines the sociodemographic profile of Spanish news consumers in the last decade. Methods: The study is based on ten surveys conducted by the Centre for Sociological Research (CIS, according to its initials in Spanish) between 2006 and 2017, on
samples ranging from 2,455 to 3,191 participants, to determine whether sociodemographic profiles of news consumers are different across media platforms and whether they can efficiently predict news consumption.
Results: It is concluded that the Internet has increased the consumption of print and digital newspapers. News consumption in radio and television is so widespread that the selected sociodemographic variables do not allow us to detect noticeable differences. However, newspaper readers are predominantly male, urban, middle-aged, middle and upper class, and highly educated. Internet news consumers basically match the previous profile, except in terms of age, as they are younger. In the case of newspaper and Internet news, the variables that predict their consumption more efficiently are education level, female sex (negative correlation) and social class (positive correlation). Finally, the correlation coefficients between news consumption in all three media platforms are significant, but not very high.

Keywords: competition, news media, news consumption, sociodemographic predictors.

Introducción: La presente investigación tiene como propósito conocer la composición sociodemográfica de los consumidores de noticias en España en la última década. Metodología: A partir de diez encuestas aplicadas por el Centro de Investigaciones Sociológicas (CIS) entre 2006 y 2017 con muestras de entre 2.455 y 3.191 personas, se buscó determinar si los perfiles sociodemográficos de consumo de medios son distintos entre sí y si predicen eficientemente el consumo.
Resultados: Se concluye que Internet ha aumentado los niveles de consumo de los diarios impresos y digitales. El consumo de noticias en radio y televisión están tan generalizado que las variables sociodemográficas elegidas no permiten establecer diferencias apreciables. Sin embargo, los lectores de periódicos son más masculinos, más urbanos, de edades intermedias, de clases medias y altas y con mayor nivel formativo que la media de la población. Los usuarios de noticias en Internet coinciden básicamente con el perfil anterior excepto en la edad, ya que su público es más joven que la media. En el caso de los periódicos e Internet, las variables que predicen su consumo más eficientemente son el nivel de estudios y el sexo femenino (en negativo) y la clase social (en positivo). Por último, los coeficientes de correlación entre el consumo de noticias en los tres medios son significativos, aunque no muy altos.

Palabras clave: competencia, intermedios, consumo de noticias, predictores sociodemográficos.

1. Introduction. 2. Methods. 3. Results. 3.1. News consumption frequencies. 3.2. Sociodemographic profile of news consumers. 3.3. Predictors of news consumption. 3.4. Correlations between news consumption in different media. 4. Discussion. 5. Conclusions. 6. References.

Juan Pablo Artero. University of Zaragoza. Spain.
Víctor Orive. University of Zaragoza. Spain.
Pilar Latorre. University of Zaragoza. Spain.

Received: 19/02/2020.
Accepted: 02/03/2020.
Published: 31/07/2020.

How to cite this article / Standard reference
Artero, J. P., Orive, V., & Latorre, P. (2020). Sociodemographic profiles and predictors of news consumers. Revista Latina de Comunicación Social, (77), 55-72.

Translation by C. A. Martínez-Arcos (PhD, University of London)

1. Introducción

In recent years, the way society consumes news media has generated great concern among media professionals and social organisations. According to a rather widespread view, the massification of the Internet as a news medium has led to a decreasing interest in traditional news media, such as newspapers, radio and television. In turn, it is increasingly argued that this decline in news consumption is caused not so much by sociodemographic factors but rather by consumers’ personal interests and lifestyles. This situation can have negative effects on the democratic system as it could lead to the formation of a less and badly informed citizenry.
Based on this context, the purpose of this research is to analyse the sociodemographic profile of news consumers in Spain in the last decade. The focus is to statistically describe consumption frequencies across different news media, as well as the sociodemographic profile of news consumers and the potential predictors of news consumption, especially structural variables, such as age, sex, residence, social class and education level. The research problem lies in the difficulty of collecting meaningful data to describe and understand a reality as complex as news consumption. In addition, previous studies have looked at specific environments and specific age groups, making it difficult to carry out overall assessments. Finally, the consolidation of the Internet as a news medium has posed conceptual and methodological challenges that are difficult to ignore.
Initially, it can be argued that news consumption on the Internet does not differ drastically from that in traditional media. However, research on this area often has three limitations: the premise of the division between print, audiovisual and digital media; the notion that the analysis should take into account the characteristics of the media and social practices separately; and the inclination to focus on patterns of either ordinary or extraordinary phenomena, but not on both at the same time (Mitchelstein and Boczkowski, 2010). When it comes to understanding how audiences see journalism today, media consumption should be conceived not only as something people do, but as something people do in a particular place. Journalism is now produced to enable increasingly mobile places of consumption (space); to adjust for the faster pace of the information age (speed); and to interact with and provide multiple access channels for audiences (convenience) (Peters, 2012).
In addition, mobile devices have caused a paradigm shift in news accessibility and consumption in daily life. Most contemporary research has been based on four thematic areas: patterns, people, places and participation (Westlund, 2015). At the same time, digital consumption in the workplace reveals some discontinuities with respect to print and audiovisual news, for example, the relevance of notions of routine, space, time, and sociability. However, it is also necessary to renew the understanding of how these concepts change when the medium changes from traditional to digital and when consumer practices coincide with work practices (Boczkowski, 2010).
Far from having lost their importance, traditional media such as newspapers and television have emerged as the most important providers of digital local news (Waldman et al., 2011). In the United States, 1,800 local newspapers have been closed since 2004, while other media such as television and digital initiatives have made efforts to fill the void left by local newspapers (Abernathy, 2018), which has also happened in Spain (Manfredi, Rojas and Herranz de la Casa, 2015). The national media survey carried out in Spain (Estudio General de Medios, 2018) has revealed that the social penetration of print newspapers has declined in the last twenty years, going from 36 to 22 per cent. Radio penetration has increased from 53% to 57%, while that of television has gone from 89% to 85%. These data alone would be devastating if not because the penetration of the Internet in the same period, 1998-2018, has gone from just 2% to 77%. While not all users consume news content every day, it could well be assumed that losses in traditional media audience could have been offset by increases on Internet use.
According to some recent contributions, most people still resort to traditional media to stay informed, although they also use online news services in certain circumstances as a supplement. People show a clear tendency to consume the brands they trust, but the news on mobile devices are gradually infiltrating the daily lives of audiences who had become disconnected from the news (Van Damme et al., 2015). It can be said that traditional and digital platforms provide the same gratifications. In fact, consumption of multimedia news does not lead to a decline in traditional news media consumption. Moreover, some research points out that television consumption reinforces consumption of digital news and vice versa. This confirms that citizens, young audiences in this study, use old and new media platforms complementarily (Van Cauwenberge, d’Haenens and Beentjes, 2010).
According to the literature, those who consumed newspapers or television news in the past are now using multiple platforms simultaneously. Alternatively, audiences can shift their attention to digital media or avoid the news at once. However, although the value of analysing specific media consumption patterns must be recognised, there is no way to measure the total consumption of news if individual consumption in different media cannot be aggregated. In fact, some segments of the epopulation use multiple sources of news and most Internet news consumers are multichannel users, i.e., they also read the newspaper and watch TV news. Consumers of online news are more likely to also consume news in traditional media than non-consumers of Internet news. In fact, there is no evidence for a decline in news consumption since the advent of the Internet, and it is likely that total news consumption across media is increasing (Ksiazek, Malthouse and Webster, 2010).
Some authors argue that the behaviour of media consumers is not very different in demographic or socioeconomic terms (Pentina and Tarafdar, 2014). However, the consumption of paid news is related to predictor variables such as age, income, purchase of other digital products and use of social networks. Similarly, those who already pay for print news are more likely to pay for digital news in the future. The intention to pay also depends on other personal factors, such as perception of value, informative interest and general media consumption patterns (Goyanes, Artero and Zapata, 2018).
Surveillance and escapism gratifications are consistent predictors of news consumption. However, habit strength is the most powerful predictor of news consumption overall. The consumption of hometown newspaper sources is strongly related to current events knowledge (Diddi and LaRose, 2006). News consumption, therefore, seems to depend more on the news category, reader gender and interest in certain topics, than on whether the news is offered in print or digital platforms. In fact, online readers do not consume and retain news differently from readers of print news media (d’Haenens, Jankowski and Heuvelman, 2004). Factors such as income level, DMA-level population density and minority levels would not affect the robustness of the news supply (Napoli et al., 2018).
When selecting their news sources, people choose those with ideological positions that are similar to their own (Chanand Suen, 2008). People want to consume news stories that are consistent with their tastes and previous beliefs. However, this fact does not necessarily harm conscientious consumers, who may be able to collect more news content from multiple media aoutlets as the latter become biased (Xiang and Sarvary, 2007). Social media increase this relationship of similarity with consumer’s own preferences, as they provide relevant news stories that are filtered by people who think similarly. However, audiences are also exposed to unverified, anonymous and more subjective sources (Pentina and Tarafdar, 2014). In general, media consumption stimulates civic and political participation. For example, in the U.S. markets where local Spanish-language television exists, Hispanic voter turnout increases by 5 to 10 percentage points (Obeholzer-Gee and Waldfogel, 2006). In general, demographic and political-social factors count towards determining voter turnout. However, voter turnout is also influenced by media consumption habits, such as listening to the radio and being interested in the news (Livingstone and Markham, 2008).
Since young adults are more intensive users of the Internet, the effects of their preference for digital media are stronger for them than for older people (Bachmann et al, 2010). Among young people, the emergence of social media as news sources, and the decline of traditional media, especially newspapers, is more evident. The appeal of free content is also greater. There is also a difference in consumption in favour of males and older people, as news consumption increases as young people mature (Casero-Ripollés, 2012). However, the news industry needs to perform a true digital convergence by providing the younger generation a multimedia news consumption experience that is customisable and relevant to them (Huang, 2009).
This brief literature review highlights some disagreement regarding the exact magnitude of the change that the Internet has produced in news consumption and the factors that are most likely to predect it. This research is based on the following research questions

The hypotheses are based on the previous literature review, which is contradictory in part, but generaly supports the aforementioned approaches, and particularly on the contributions of Livingstone and Markham (2008), Ksiazek, Malthouse and Webster (2010), Mitchelstein and Boczkowski (2010), Van Cauwenberge, d’Haenens and Beentjes (2010), Waldman et al., (2011), Casero-Ripollés (2012), Abernathy (2018) and Goyanes, Artero and Zapata (2018).

2. Methods

The survey was chosen as the method to develop this article, although using secondary data collected over the years by the Spanish Centre for Sociological Research (CIS). To be preceise, the monthly surveys carried out periodically by the CIS include, at least since 2000, a question about the frequency of news consumption in different media. Based on the results of eight of these surveys, conducted from 2000 to 2009, Meilán (2010) developed his research work, which this article partially tries to replicate. Thus, the results obtained in both works are compared in the discussion section. However, it must be noted that Meilán combined different question and answers criteria. The first two years, the survey used a six-point consumption scale, while from 2004 onwards it used the classic Likert five-point scale. Moreover, the first four surveys, carried out between 2000 and 2005, measured the consumption of radio and television separately. These differences in the survey questionnaire undoubtedly made it difficult to obtain consistent and reliable data throughout the period of study.
To address this methodological issue, this study only takes into account those CIS surveys that included the exact same question and answer options. After a brief analysis, it became apparent that the question model A was available for ten years, from 2006 to 2017 (except for 2009 and 2010), while other two slightly different versions were available in four years each. These last surveys were dismissed and a database was built with the data provided by the CIS for the ten years mentioned above, which correspond to the following surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191. The analysis focused on the answers to three questions: “Do you read the political sections of the newspaper?”, “Do you consume news on radio and television?” and “Do you use the Internet to get the news about politics and society?”. A fourth question was dismissed because it was redundant and did not add relevant information. The dissmissed question is: “Apart from the news, do you consume other programmes about politics on radio and televisión?”
The selected surveys carried out by the CIS are telephone surveys applied to a representative sample of the Spanish population over the age of 18. The number of interviews conducted in each of the surveys varies, ranging from 2,455 in 2007 to 3,191 in 2006, as it will be shownone, as evidenced in Tables 4, 5 and 6.  the arrays has been SPSS 22.0.
xcepto 2009 and 2010). They're unlocked. Spicein the following tables. The software used for the analysis of the survey data sets is SPSS 22.0.
In the subsequent data processing, consumption frequency variables were recoded in the following way: “every day” was assigned five points and “never”, one point, as shown in Tables 4, 5 and 6. The sociodemographic variables that were correlated with the answers are age, sex, residence, social class and education level, as well as ideology (ranging from 1 for left-wing to 10 for right-wing). Similarly, the education level variable included in the sociodemographic questions section at the end of the questionnaire offered the option “university media” from 2006 to 2013, but not thereafter, which has not posed an insurmountable difficulty. Finally, to create tables 7, 8 and 9, the options “male” and “female” were recoded with the values 0 and 1, respectively, while the categories urban and rural residence (for municipalities with more and less than 50,000 inhabitants, respectively), which were recoded as 1 and 0, respectively.
The selection of the CIS survey database for this research is justified by its undoubted professionalism, availability of technical and human resources and the ability for extracting data sets over a large period, which can yield more robust results than the application of a single survey today. These historical data allow us to examine the evolution of online news consumption as it was reported back then, and not as a retrospective survey. The rigour of the data collection and processing is also proven, ensuring that the matrices are scientifically reliable.

3. Results

3.1. News consumption frequencies

Table 1. Frequency of newspaper political news consumption.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

The proportion of the population who read newspapers daily have increased between 2006 and 2017 from 16% to 23%, albeit to the slight detriment of those who read it less frequently. This is confirmed by the fact that the percentage of people who never read newspapers has remained stable, between 36% and 38%. In other words, newspaper reading has not declined in the decade under study. This percentage would not be consistent with other known data if it only referred to print newspaper consumption, but it is clear that surveyed participants referred to both print and digital newspapers.

Table 2. Frequency of radio and television news consumption.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

However, the percentage of Spaniards who consume radio and television news on a daily basis is much higher (two-thirds) and has remained stable from 2006 to 2017. Intermediate frequencies are similar in the 2006-2017 surveys, while the percentage of those who never consume audiovisual news has increased slightly: from 3% to 6%.

Table 3. Frequency of political and social news search online.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

However, the frequency of search for political or social news on the Internet has increased significantly between 2006 and 2017. The percentage of those who search for online news everyday has increased fivefold: from 4% to 25%. In addition, the percentage of those who do so between one and four days a week has increased from 3% to 8%. Finally, the percentage of the population who does not consume Internet news has fallen almost by half, from 78% to 46%.

3.2. Sociodemographic profile of news consumers

Table 4. Sociodemographic profile of readers of political sections in newspapers.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

Newspaper readers are predominantly male and urban consistently throughout the series. In terms of age, readers aged 34 to 65 read more frequently than younger and older readers. With regards to their social class, the higher it is, the more frequently they consume the press. The same goes for the level of studies: the more years of education readers have, the more frequently they read newspapers. Finally, the more left-wing respondents are, the more frequently they read newpapers, although the differences are little significant. Similarly, in the series of surveys observed, the changes detected throughout the decade are minor.

Table 5. Sociodemographic profile of radio and television news consumers.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

With regards to the consumption frequency of audiovisual news, the sociodemographic variables under analysis do not allow for significant differences. That is, all groups of respondents consume audiovisual news very often regardless of their sex, age, municipality, social class, level of studies and ideology. This situation has hardly changed in the decade under study.

Table 6. Sociodemographic profile of Internet news consumers.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

In terms of the search for political or social news on the Internet, male and urban users do so more frequently. As for the age variable, the correlation is inverse: the older the user, the lower the frequency of digital news serach. In terms of social class, the higher the class of the user, the higher the search frequency. The same goes for the educational level. Overall, online news search frequency has almost doubled, except among people with non-formal education or only primary education.

3.3. Predictors of news consumption

Table 7. Predictors of newspaper political content consumption.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

Adjusted R-Squared indicates the extent to what the six independent variables predict the dependent variable. Sex, residence, social class, level of studies and political ideology can explain between 17% and 22% of newspaper consumption over the years.
Predictive variables have a low significance level, of less than 0.05. According to the results, the following variables are excluded: old middle classes and skilled workers (throughout the ten years), higher education (in eight years) and secondary education (in six years). It is also necessary to take with caution the following variables: vocational training (in four years), non-formal education (in three years), primary education (in three years), municipality type (in three years), new middle class (in three years) and upper class (in one year).
The best predictors of newspaper political content consumption are non-formal education, primary education, vocational training, sex (all of them negative) and upper/upper middle class and new middle classes (positive). In other words, the lower the education level, the lower newspaper consumption is, as it is the case with women. And the higher the social class, the higher newspaper consumption is.

Table 8. Predictors of news consumption in radio and television.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

With regards to audiovisual news consumption, R-Squared indicates that the variables chosen are only capable of explaining between 3 and 9 per cent of it. Most variables are not significant. The only variable with a high negative coefficient is non-formal education, while age has a low positive coefficient. That is: the lower the education level, the lower the consumption of radio and television news. And the older the respondent, the more news programmes he or she consumes.

Table 9. Predictors of online political and social news consumption.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

With regards to the consumption of online news, R-squared indicates that the chosen variables explain between 9 and 25 per cent of it, with a clear upward trend. However, some variables must be ruled out because they are not significant in most years. Specifically: all social classes, except for the upper/upper middle class, and all education levels, except for non-formal education, primary education and vocational training.
The most efficient predictors of Internet news consumption are non-formal education, primary education and vocational training, all three with negative coefficients. However, it should be noted that in the last three years under study, the figures are not significant. The sex variable also has negative and significant coefficients in all years but one. On the other hand, the upper/upper middle-class variable has yielded high positive coefficients, especially in recent years. This means that the lower the education level, the lower the online news consumption. Women also use the Internet less for this purpose. And finally, the higher the social class, the higher the consumption of online news.

3.4. Correlations between news consumption in different media

Table 10. Pearson’s correlation coefficients between different media.

Source: authors’ own creation with data from the following CIS surveys: 2632, 2700, 2749, 2914, 2960, 3001, 3041, 3114, 3156 and 3191.

With regards to the correlations of consumption between the three media platforms, there are several situations. The Pearson’s correlation coefficient ranges from -1 to1. When the Pearson’s correlation coefficient is between 0.7 and 1, the correlation is strong and direct; when it is between -0.1 and 0.1, the correlation is very weak; when it is 0, there is no correlation between variables; and when it is between -0.7 and -1, the correlation is strong and inverse. This does not imply causation, but only linear relations between two variables.
In these data sets, the correlation coefficients are significant and indicate that people consume news in different media. However, they are not very high coefficients. Newspaper readers have the highest correlation coefficients in comparison with other media users, both radio and television (always around 0.2) and, above all, Internet (between 0.2 and 0.4). The trend is to increaser over the years. In this way, the correlation between the consumption of newpaper and Internet news is the highest in the entire data sets series, although the correlation level is medium.

4. Discussion

In Meilán (2010), the percentage of daily newspaper readers between 2002 and 2004 fell significantly (37%) but remained stable since then and until 2010. In 2011, the number of daily newspaper readers increased, while the number of people who reads the newspaper less frequently declined. Thus, people who never read about politics in newspapers are always about a third, which remains fairly stable between 2000 and 2017. On the other hand, radio and television are the dominant news media sources now, with about two-thirds of respondents reporting news consumption in these media every day or almost every day. In the 2000-2009 and 2006-2017 periods, the figures are therefore very similar. Finally, daily consumption of online news increased between 2004 and 2009, going from 3% to 11%. This sustained growth rate reached 25% in 2017.
Meilán (2010) suggests that fluctuations in media consumption may be due to different levels of interest in news across the years. However, the evidence for the Spanish case is weak. It is necessary to find explanations in technological changes. From this point of view, the press, radio and television underwent slower changes as well as their consumption. Television experienced an increase in the number of channels between 2006 and 2017, but consumption did not rise. In the press, the only major change was the emergence of free newspapers, but their boom was very brief and did not significantly affect the overall consumption. However, the increasing penetration of the Internet in Spanish households had an impact in the search for online political and social news in a very prominent way.
As for the sociodemographic profile of news consumers in different media, in the 2000-2008 period, most readers were male, with higher consumption among high-schoolers and university students, and between 35 and 49 years of age, while consumers over 65 were well below average. Moreover, upper-class respondents read the press more than the average. All these distributions were also confirmed between 2006 and 2017, although the reader base increased above average between the ages of 35 and 64.
Radio and television data are presented jointly only in Meilán (2010) between 2006 and 2009, so it is not possible to draw relevant comparisons with the full period of this study (2006-2017), which already includes those years. In the case of the Internet, previous research found that it was the medium where sociodemographic boundaries were sharper. Between 2004 and 2009, males, high school students, college students and people under 35 searched for online news more than the rest of the population. Age was the key variable, with higher consumption in the 18-34 group and a much lower consumption in people 65 and over. The same occurred with the two highest social classes. Data between 2006 and 2017 confirm these trends. Men effectively consume more online news than women. Age now widens to establish a completely inverse correlation in groups: the older the person, the lower his of her consumption of online news. The same goes for social class and education level, where the relationship is direct. The big difference is that consumption frequencies have doubled between 2006 and 2017, even if that is not the case among people with non-formal education or primary education.
Overall, the consumption of political news in large cities was always higher in all three media, although on radio and television the differences were minor. Thus, the rural gap was evident in the press and the Internet, but not in audiovisual news. Respondents’ ideological self-placement had no major differences, although there was a slight increase in press and online news consumption among left-leaning respondents and in radio and television news consumption among right-leaning respondents. Between 2006 and 2017 these trends were confirmed.
In conclusion, Meilán’s study (2010) found marked sociodemographic inequalities in news consumption. Males, middle-aged people, with middle and high education levels, and residents of urban centres, had greater access to news. While women, people in lower classes and with low education levels reported lower news consumption altogether. Differences were more marked in the press and the Internet than in radio and television, suggesting a differentiated access to different media. These conclusions remain valid for the 2006-2017 period.
In 2010, predictors explained between 23% and 38% of news consumption in the press. This research predictors only reach between 17% and 22%, as the variables employment situation, interest in politics and consumption of other media have now been excluded. Therefore, it was shown that at an older age and higher educational level, there is a higher consumption. Women and lower social classes read newspapers to a lesser degree. Age and urban residence were uncovered as emerging predictors. In the 2006-2017 period, good predictors are middle and upper classes (positive), low education levels and the female sex (negative). However, the influence of the variables age, residence and ideology does not seem to be relevant.
With regards to radio and television news, between 2006 and 2009, the variables explained only 4 to 8 per cent of the consumption and between 2006 and 2017 the percentages are similar (between 3 and 9 per cent). Thus, only age was a positive explanatory factor, albeit with a low coefficient. This variable behaves in the same way again. The novelty now is that having non-formal education has a high negative coefficient. That is: the older the person, the greater the consumption of news; and the lower the education level, the less interest is shown in radio and television news. The model does not work properly for these media.
Regarding the consumption of political and social news on the Internet, between 2004 and 2009 the variables predicted between 14 and 21 per cent of the consumption. Between 2006 and 2017 the range changed to between 9 and 25 per cent, with an upward trend. This means that sociodemographic variables increasingly explain the consumption of digital news. In the first study, all social class categories predicted below-average consumption. Meanwhile, age and education level were positive predictors as well as the female sex category. Between 2006 and 2017, the most efficient predictors were: non-formal education, primary education and vocational training (in negative), upper and upper middle class and female sex. That is: the lower the education level, the lower the consumption of online news; the higher the social class, the higher the consumption of online news; and women tend to consume less digital news than men do.
Finally, the crossing of the consumption variables among the three media platforms demonstrated a high correlation between 2000 and 2009. Therefore, the higher the consumption of news in the press, radio, television and the Internet, the greater consumption in any of the other media and vice versa as consumption was lower. Higher coefficients corresponded to the press in comparison to the rest of the media. These correlations have been fully confirmed with similar indicators between 2006 and 2017.

5. Conclusions

The problem of study in this article is based on the relative discrepancy in the literature about how the Internet has affected the consumption of news in traditional media and the possible validity of sociodemographic criteria as predictors of news consumption. The main conclusions of this study are set out below, according to the research questions presented above.

RQ1) Has the Internet brought about a significant change in the levels of news consumption in traditional media?

The number of daily newspaper readers, of both print and digital publications, has increased from 16 to 23 per cent. In the case of radio and television news, their daily consumption has remained stable at around 67 per cent. The Internet has experienced a large increase in terms of users who search for news daily, going from 4 to 25 per cent. Thus, the change has been on the rise for both digital newspapers and the search for political or social news on the Internet.
However, the data of non news-consumers has changed very slightly. The percentage of people who do not read the press has remained stable over the decade, ranging from 36 to 38 per cent, while the percentage of those who never consume radio and television news ranges from 3 to 6 per cent; and the proportion of those who never search for online news has declined from 78 to 46 per cent.

RQ2) Is the sociodemographic profile of news consumers very different depending on the medium in question?

The data show that the consumption of news in radio and television is so widespread that the chosen sociodemographic variables do not allow for noticeable differences. However, in the case of newspaper readers, they are mostly male, urban, middle aged (35-64), middle and upper class and with a high level of training. Internet news users basically match the previous profile except for age, as they are younger than the average news consumer. Thus, there are two large groups of news consumers: those who consume news on radio and television, who are similar to the national average; and those who read news in newspapers and on the Internet, who exhibit traits that are different than the average but are common to each other except for age.

RQ3) Are sociodemographic variables still good predictors for news consumption?

The six variables chosen (including ideological self-placement) are capable of explaining between 17 and 22 per cent of newspaper consumption, between 3 and 9 per cent of audiovisual news consumption, and between 9 and 25 per cent of internet news search, depending on the year. Therefore, their predictive capacity for the consumption of audiovisual news is very low. In the case of newspaper and the Internet news, the variables that most efficiently predict their consumption are the level of studies and the female sex (negatively) and social class (positively). Therefore, three of the six variables chosen, in two of the three news media platforms can be considered good predictors.

RQ4) Does news consumption in one medium significantly predict consumption in other media?

The correlation coefficients between news consumption in all three media are significant, although they are not very high. Newspaper readers have the highest coefficients with the other two media.
These results may be considered satisfactory but do not hide the limitations of the study. The question and answer options included in the survey questionnaire could have been formulated more clearly, to define political and general news and the appropriate separation of print and digital media, radio and television. This was not possible to do in this article because of the secondary information available. In addition, many other sociodemographic and psychographic variables could have been measured. However, the CIS’s surveys, of course, have introduced their own criteria for the extension of the questionnaire. Finally, audience and reception studies in general continue to need basic research and intermediate research that pays more attention to time as a unit of measure for any act of media consumption. When academic and commercial research becomes able to collect this type of data, new and interesting lines of research could be developed around the consumption of news content, entertainment and marketing in the media.


  1. Abernathy, P. M. (2018). The expanding news desert. Center for Innovation and Sustainability in Local Media. University of North Carolina at Chapel Hill.
  2. Asociación para la Investigación de los Medios de Comunicación (AIMC) (2018). Resumen general del Estudio General de Medios (EGM): año móvil febrero a noviembre 2018. Madrid.
  3. Bachmann, I., K. Kaufhold, S. Lewis, & Gil de Zúñiga, H. (2010). News Platform Preference: Advancing the Effects of Age and Media Consumption on Political Participation. International Journal of Internet Science, 5(1), 34-47.
  4. Boczkowski, P. J. (2010). The consumption of online news at work: making sense of emerging phenomena and rethinking existing concepts. Information, Communication & Society, 13(4), 470-484.
  5. Casero-Ripollés, A. (2012). Beyond Newspapers: News Consumption among Young People in the Digital Era. Comunicar, 20(39), 151-158.
  6. Chan, J., & Suen, W. (2008).  A Spatial Theory of News Consumption and Electoral Competition. The Review of Economic Studies, 75(3), 699-728.
  7. D’Haenens, L., Jankowski, N., & Heuvelman, A. (2004). News in online and print newspapers: differences in reader consumption and recall. New Media & Society, 6(3), 363-382.
  8. Diddi, A., & LaRose, R. (2006). Getting Hooked on News: Uses and Gratifications and the Formation of News Habits Among College Students in an Internet Environment. Journal of Broadcasting & Electronic Media, 50(2), 193-210.
  9. Goyanes, M., Artero, J. P., & Zapata, L. (2018). The effects of news authorship, exclusiveness and media type in readers’ paying intent for online news: An experimental study. Journalism.
  10. Huang, E. (2009). The Causes of Youths’ Low News Consumption and Strategies for Making Youths Happy News Consumers. Convergence: The International Journal of Research into New Media Technologies, 15(1), 105-122.
  11. Ksiazek, T. B., Malthouse, E. C., & Webster, J. G. (2010). News-seekers and Avoiders: Exploring Patterns of Total News Consumption Across Media and the Relationship to Civic Participation. Journal of Broadcasting & Electronic Media, 54(4), 551-568.
  12. Livingstone, S., & Markham, T. (2008). The Contribution of Media Consumption to Civic Participation. The British Journal of Sociology, 59(2), 351-371.
  13. Manfredi Sánchez, J. L., Rojas Torrijos, J. L. y Herranz de la Casa, J. M. (2015). Periodismo emprendedor: el periodismo deportivo en España. Revista Latina de Comunicación Social, (70), 69-90.
  14. Meilán, X. (2010). Causas y consecuencias del consumo de información política en España (2000-2009). Colección Opiniones y Actitudes, 68. Centro de Investigaciones Sociológicas.
  15. Mitchelstein, E., & Boczkowski, P. J. (2010). Online News Consumption Research: An Assessment of Past Work and an Agenda for the Future. New Media & Society, 12(7), 1085-1102.
  16. Napoli, P., Weber, M., McCollough, K., &  Wang, Q. (2018). Assessing Local Journalism: News Deserts, Journalism Divides, and the Determinants of the Robustness of Local News. DeWitt Wallace Center for Media and Democracy, Duke University.
  17. Obeholzer-Gee, F., & Waldfogel, J. (2006). Media Markets and Localism: Does Local News en Español Boost Hispanic Voter Turnout? National Bureau of Economic Research Working Paper 12317.
  18. Pentina, I., & Tarafdar, M. (2014). From ‘‘information’’ to ‘‘knowing’’: Exploring the role of social media in contemporary news consumption”. Computers in Human Behavior, (35), 211-223.
  19. Peters, C. (2012). Journalism to go. Journalism Studies, 13(5-6), 695-705.
  20. Van Cauwenberge, A., d’Haenens, L. & Beentjes, H. (2010). Emerging Consumption Patterns among Young People of Traditional and Internet News Platforms in the Low Countries. Observatorio (OBS*) Journal, 4(3), 335-352.
  21. Van Damme, K.,  Courtois, C., Verbrugge, K., & De Marez, L. (2015). What’s APPening to news? A mixed-method audience centred study on mobile news consumption. Mobile Media & Communication, 3(2), 196-213.
  22. Waldman, S., and the Working Group on Information Needs of Communities (2011). The information needs of communities. Federal Communications Commission.
  23. Westlund, O. (2015). News consumption in an age of mobile media: Patterns, people, place, and participation”. Mobile Media & Communication, 3(2), 151-159.
  24. Xiang, Y., & Sarvary, M. (2007). News Consumption and Media Bias. Marketing Science, 25(5), 611-628.


Juan Pablo Artero
Juan Pablo Artero is PhD in Communication and Associate Professor of Journalism at the University of Zaragoza. Since March 2018, Vice-chairman of the board of directors of the Aragonese Radio and Television Corporation (CARTV).
Índice H: 8.
Orcid ID:
Google Scholar:

Víctor Orive
PhD in Economics and Business Organization and assistant professor of PhD at the University School of Tourism of the University of Zaragoza. He teaches Market Research and Marketing and quality of tourist destinations.
Índice H: 3.
Orcid ID:
Google Scholar:

Pilar Latorre
Pilar Latorre is an industrial engineer, PhD in Economics and Business Administration and Assistant Professor of Business Administration at the University of Zaragoza. Her research work focuses on operations, performance evaluation and benchmarking.
Índice H: 5.
Orcid ID:
Google Scholar: