A case of reverse-agenda setting? How 2018’s FIFA World Cup coverage reduced media reporting of Uruguayan budget bill’s yearly revision

By Matías Dodel, Federico Comesaña, and Daniel Blanc

[Citation: Dodel, M., Comesaña, F., & Blanc, D. (2019). A case of reverse-agenda setting? How 2018’s FIFA World Cup coverage reduced media reporting of Uruguayan budget bill’s yearly revision. #ISOJ Journal, 9(1), 13-28.]

Through agenda setting, news media become critical for the visibility of political account­ability instances. This article aims to provide statistical evidence for a scenario in which news media shift their reporting agenda anticipating the public’s interests or newswor­thiness of an extraneous event (the 2018 FIFA World Cup), reducing the coverage of a  critical one-time-a-year accountability instance (the “Rendición de cuentas”). Based on scrapping all publications from the main Uruguayan news media conglomerates a lagged  dependent variable Poisson model was fitted on the “Rendición de cuentas” daily news reporting. Findings signal the necessity to consider the externalities of conducting critical democratic debates during mega-events.

 

News media reporting plays several critical roles in modern liberal democracies (Law­rence & Suddaby, 2006). This article will focus only on one of these: their potential to modify the exposure, societal discussion and debate of diverse policy issues.

In a perfect-world scenario, citizens are aware and monitoring all critical topics plau­sible to affect their lives. Nonetheless, in reality a plethora of issues compete for a finite agenda space, which may shift public’s attention or news media interests away (Dader, 1990).

This article studies one of these cases, by assessing the effects of an extraneous sport event (the 2018 FIFA World Cup) on the report of a critical one-time-a-year national accountability event in Uruguay: the Uruguayan budget bill’s yearly revision, known as “Rendición de cuentas” (see Moraes, Chasquetti & Bergara, 2005 for further details on how the bill’s yearly revision fit the Budgetary Process of the Public Sector).

Whereas there is an important cluster of literature on how media reporting sets public agendas on a plethora of issues (McCombs & Valenzuela, 2007), there is consistently less non-anecdotal evidence on news media shifting their reporting agendas anticipating the public’s interests (Russel et al., 2014). This article aims to contribute to the literature by providing empirical evidence of a case of this phenomenon.

The article proposes that the 2018’s FIFA World Cup produced a change in the trends of the reporting of the Uruguayan budget bill’s yearly revision. In order to provide evidence for this, we developed software crawlers which scrapped all publications within the web-sites of the three biggest Uruguayan news media conglomerates (El País, El Observador & Montevideo Portal) for an almost two-year period. We then identified all publications concerning the budget bill’s yearly revision in order to analyze and fit a lagged depen­dent variable Poisson model on the count of “Rendición de cuentas’” daily news. We compare the budget bill’s yearly revision reporting with and without the effect of the World Cup near its parliamentary entry, providing statistical evidence for a significant decrease of the daily reporting during the 2018 World Cup.

Agenda Setting

Rogers and Dearing (1988) identify two research traditions concerning the study of agenda setting: public agenda setting by the mass media, and policy agenda setting. Rogers and Dearing (1988) postulate that both traditions are nurtured by Cohen’s 1963 seminal work, but whereas the latter started earlier, the former resulted more prolific.

The mass media and public agenda setting’s literature conceives agenda setting as the effect of media reporting by setting public agendas more than public opinions itself; “telling people what to think about” not “what to think” (Cohen, 1963; McCombs & Shaw, 1972; Russel et al., 2014). In the words of Dader (1990, p. 295), the agenda setting ef­fect is “… nothing but the orientation, conduction or channeling that the minds of citizens suffer towards a repertoire of issues of public concern, to the detriment of others that are not mentioned or highlighted.” Exposure over time of a reduced number of issues or people, makes them prevalent in the minds of individuals and public opinion (McCombs & Valenzuela, 2007).

In the words of some of the fathers of the agenda setting theory, the core concepts of are: an object agenda, and attribute agenda, and the transfer of salience between pairs of agendas (McCombs, Shaw, & Weaver, 2014, p.782). These authors (McCombs et al., 2014) identify seven facets of the development of the agenda-setting theory in its first 50 years: a) the first level of agenda setting (i.e. the salience of objects); b) a second level, concerning the attributes of the salient objects (see Kim, Scheufele & Shanahan, 2002; McCombs & Valenzuela, 2007); c) a third level of agenda setting focalizing in the net­worked impact of previous levels; d) psychological bases of agenda setting and the no­tion of need for orientation; e) the consequences of agenda setting for attitudes, opinions and behaviors; f) the merging of media, public and personal views to create satisfying pictures of the world; and g) the origins and determinants of media agendas.

Our article aims to contribute to the understanding of the latter only, the supply side of the agendas we may add, by assessing the peculiar scenario previously described. McCombs et al. (2014, p. 782) identify several factors affecting media agendas such as the “… prevailing cultural and ideological environment to news sources …”, the influence and relationships within the media, the norms and routines of journalism, or the individual characteristics of journalists.

Our study aims to provide evidence on the plausibility of perceived newsworthiness, or news media shifting their reporting agendas anticipating the public’s interests (Russel et al., 2014), as another determinant of media agenda.

Reverse-agenda setting and digital technologies

The traditional understanding of agenda setting proposes that, under finite time and space to cover an almost unlimited number of issues and events, news media make some decision and focus their attention only on some topics (Dader, 1990, p. 295). Thus, the relationship’s direction goes from media to the public, varying across issues and au­diences, in the strengths and lags of media effects over public opinion (McCombs, 2005; McCombs & Shaw, 1993).

While the correlations between media and public agendas has already been attested, besides scarce anecdotal evidence, the inverse causal direction of this relationship has been far less studied, and results in more ambiguous findings (Aruguete, 2017; Russel et al., 2014).

With the advent of the Internet and online news media reporting, two innovations en­abled the improvement of the study of this alternative path of effects. On one side, digital technologies allow researchers to assess almost in-real-time influence or contagion of social media trends to news media reporting and vice versa (Russel et al., 2014). Russel et al. (2014) provided empirical evidence that agenda setting does not follow a one-way pattern from traditional media to individuals, at least concerning political issues in the United States. Aruguete (2017) reviewed the literature that discusses theoretical and empirical consequences of Agenda Setting theory in the relationship between traditional media and new media, broadly corroborating Russel et al.’s (2014) findings in wider contexts. Nonetheless, as most studies have been purely correlational, Aruguete (2017) warns that finding content of one agenda within another agenda does not yield informa­tion about its potential to influence.

The second innovation provided by digital technologies refers to the potential of “big data” techniques which contribute to improve the accuracy of time stamping, text analy­sis and time analyses for two or more issues competing for new media attention at the same period. This innovation guides the research design of this article.

Political accountability, transparency and news reporting

Media are not the only source of information about public issues, as the relationship with families, friends and colleagues also informs one’s world. Thus, some issues are present in the daily life of individuals while others, mostly introduced by the media (McCombs et al., 2014). As a result, the less present a certain topic is in our daily lives, the more one will turn to the media as a source of information (McCombs & Valenzuela, 2007). In the case studied in this article, the role of media as the main information source to common citizens (i.e. not politically militant) seems to be even more prevalent, as the budget bill’s yearly revision is a political accountability and/or public transparency issue seldom to be discussed in private or family settings.

At its most basic and general level, accountability can be defined as an act or situation in which actors and/or institutions explain or divulge certain actions or situations to a forum or public: an exercise of public transparency. Political accountability is publicly relevant as it can trigger diverse responses from the public sphere and have relevant conse­quences for democratic nations (Strömberg, 2015).

Nonetheless, political actors are more prone to communicate by themselves the good news than the bad news; this is why objective or transparency political accountability instances are critical. The literature signals that responses to real accountability exer­cises are of concern for politics, because they may derive in sanctions ranging from the general disapproval of the population to the need to withdraw from the public sphere (Jacobs & Schillemans, 2016).

The media are characterized as having a specific role to play in this arena. Some au­thors even argue that accountability is one of the most significant social functions of me­dia (Lawrence & Suddaby, 2006). For example, Strömberg (2015) proposes that when the media interact with greater insistence on the elaboration of policies and generate a logic of permanent scrutiny, it is healthier for the democratic system since more informa­tion is transmitted to citizens.

Nevertheless, the relevance of the media with respect to accountability has been viewed differently within the literature. While some consider that its role is increasing both in size and specificity (Bovens, 2007), others have argued that there is no clear evidence sup­porting the claim (Monika et al, 2014). This work does not discuss these issues, but may inform regarding some risks for the accountability and transparency reporting agenda in the form of competing agenda topics or issues.

News media report, “fútbol” and national identity in Uruguay

The effect of mega-sports events on national interests is not new within the literature (Evans & Kelley, 2002; Gift & Miner, 2017; Van Hilvoorde & Elling & Stokvis, 2010), particularly for South American countries in which “fútbol” (soccer) has a strong cultural relevance at national levels (Amoedo & Queirolo, 2013; Mitchelstein et al., 2017).

Fútbol is a particularly sensitive topic for Uruguayans, partly due to two historical World Cup victories in 1930 and 1950 serving as the basis of Uruguayans’ national identity and the narrative of its modern heroes’ myths (Giulianotti, 1999; Maneiro & Marchi, 2015). As the 1950’s national team manager is credited to have said, “Other countries have their history, Uruguay has its football” (as quoted in Giulianotti, 1999).

Under a World Cup scenario in which the national team had perceived chances to win/advance to the final rounds (Uruguay advanced to quarter-finals, keeping the national team in competition for three of the four weeks of the tournament), it is feasible that journalists and/or new media’s editors could have shifted the agenda prioritizing FIFA’s World Cup by “anticipating or estimating public interests,” a premise deemed implausible by McCombs and Shaw’s original agenda setting theory (Russel et al., 2014 p. 134).

Research Hypotheses

The research objectives that guide this study are concise: to provide empirical evidence for a reverse agenda setting-like effect. Consequently, the article presents only two hypotheses to be tested. The study’s main hypothesis (H1) proposes that daily count of “Rendición de cuentas” reports will be significantly reduced during 2018’s World Cup compared to the same period in 2017 (without the mega-event).

A second hypothesis (H2) refers to the relevance of the budget bill’s yearly revision entry to the parliament for the “Rendición de cuentas” reporting. The researchers expect an increase in the trend of the daily count of “Rendición de cuentas” reporting towards its entry to the parliament (H2a), as well as a decrease in this number after the bill enters to the legislative body (H2b).

Method

Data collection

Using data mining techniques, the researchers developed software crawlers that scraped all publications within the website of three of the four biggest Uruguayan news media conglomerates: El País (www.elpais.com.uy), El Observador (www.elobservador.com. uy) and Montevideo Portal (www.montevideo.com.uy). The researchers created this list based on the IAB 2016 Report’s ranking of unique browsers entries to Internet media companies in Uruguay (IAB Uruguay, 2016). The other member of this group, Subrayado (www.subrayado.com.uy), was omitted from the data collection because it publishes content predominantly in video format, functioning as the web portal of one of the main broadcast news shows in the Uruguayan open air TV (Subrayado on Canal 10).

All text and images from the websites were scrapped, developing three different algorithms for the task, based on the “R” coding language and using the library Rvest. Besides data for news items as a whole, based on Boolean search terms we identified all publications concerning budget bill’s yearly revision.

Design

The study’s research design takes advantage of a quasi-natural experiment. The budget bill’s yearly revision coincides by mere chance and only sporadically with FIFA’s World Cup, held every four years. In other terms, every four or more years an agenda space dispute occurs between sports (World Cup) and government transparency issues (bud­get bill’s yearly revision), where neither government nor news media are able to modify the time and length of the two competing events. Consequently, comparing the budget bill’s yearly revision reporting of two years, one with and another without the effect of the World Cup, will enable to test for this reverse agenda setting effect.

In order to contrast this hypothesis, we fit a lagged dependent variable Poisson model on the count of “Rendición de cuentas’” daily news items. (1)

Additional decisions were taken to limit the model’s time scope. Because it is probable that several factors affect the news cycle far away from the budget bill’s yearly revision entry to the parliament (see Figure 1), we chose to focus exclusively on a time frame close to the parliamentary entry: approximately two months before, and two months after the entry, in each of the two years for which we collected data (2017 and 2018).

Thus, the strategy consists of two steps. As a first step, a test of identical models for two different series (2017 and 2018) were conducted in order to achieve reasonable levels of fit, controlling/correcting for time-series parameter violations, as well as to test H2a and H2b related to the trend of the budget bill’s yearly revision entry to the parliament. These models do not include any reference to the World Cup.

In a second step, a nested model for the 2018 series only was conducted, including the World Cup period as an additional explanatory variable in the nested lagged dependent variable Poisson model. If the hypothesis is correct, the second 2018 model would im­prove the first 2018’s model fit as well as its explanatory power.

Measures

Dependent variable

Count of “Rendición de cuentas” daily news items

The dependent variable is the count of all daily “Rendición de cuentas” news items. Whereas democratic accountability can be understood as a theme under the agenda setting literature, the researchers opted to work with the Uruguayan budget bill’s yearly revision entry to parliament, understanding it as an event that may trigger discussions about governmental accountability and budget distribution (Dader, 1990, p. 302).

Using boolean search terms, the researchers identified all publications concerning the budget bill’s yearly revision in each of the three news sites. This was further refined by eliminating potential news items referring to other types of accountability processes. As in Spanish the term “Rendición de cuentas” can be used also to denote any type of accountability or evaluation-like instance, the researchers looked case-by-case into uncommon news categories in which the term appeared, such as sports and show busi­ness. If the use of the term was related to the budget bill’s yearly revision, the news item was kept. If not, the entry was deleted and the variable was refined.

Independent variables

The selected independent variables refer to the World Cup time period, a dummy vari­able for the days during which the 2018’s FIFA World Cup was taking place; the days to budget bill’s yearly revision entry to the parliament, a negative integer variable which increases its value the closer to the date in which the budget bill’s yearly revision entered to the parliament in each year (value = 0); and the days since budget bill’s yearly revision entry to the parliament, a positive integer variable which increases its value the further to the date in which the budget bill’s yearly revision entered to the parliament in each year (value = 0).

Proxy variables

Additionally, two other variables common to time series analyses were used in the mod­els: a one-order lag of the dependent variable in order to get rid of the autocorrelation, and dummy variables indicating the day of the week to correct for news cycle-related seasonality issues (Keele & Kelly, 2006; Wooldridge, 2010).

Results

Table 1 shows a summary of the news items published by each of the news media websites during each year until September 8, 2018, the last available data point in the analysis (analyses were conducted during September and October 2018).

As the table shows, El País has a significantly higher count of news items as a whole, as well as “Rendición de cuentas” news. El Observador has a lower rate of daily news items of the three, but has more news items on “Rendición de cuentas” news items than Montevideo Portal. While we will not address this issue here, we hypothesize these dif­ferences in favor of El País and El Observador in relation to the count of the budget bill’s years revision news items have to do with their origins and present status as traditional news media outlets (their parent companies still produce printed newspapers).

Additionally, Table 1 shows that whereas the count of news items increased in the three news media sites and as a whole between 2017 and 2018, the reporting of “Rendición de Cuentas” was stable as a whole in the two years, decreased for El País and El Observador, and increased for Montevideo Portal.

Figure 1 shows the timeline graph of the count of “Rendición de cuentas” daily news items across the whole period for which the data was collected: January 1, 2017, to Sep­tember 8, 2017. The figure shows the World Cup period as well as the dates in which the budget bill’s yearly revision entered the parliament in 2017 (June 20) and 2018 (June, 30).

Whereas the graph has considerable noise due to the daily reporting variance, it allows one to see a positive trend in relation to “Rendición de cuentas”’ before its entry to the parliament (see the trend line).

 

Figure 2 presents the same data but zooming in on the two months prior to and after the “Rendición de cuentas” entry to parliament. Moreover, the figure compares 2017 and 2018 using the days to and from “Rendición de cuentas” entry to parliament to create the time series (zero in 2017 refers to June 20, and in 2018 to June, 30). The World Cup period is signaled in the figure. Figure 2 shows that while “Rendición de cuentas” news reporting follows a similar pattern for both years, as H2 predicted, 2018’s count appears to be higher before the World Cup period compared to 2017’s, but not during the event. Nonetheless, the relations are not completely clear and visual analyses are insufficient to assess our hypotheses with statistical certainty.

Predicting “Rendición de cuentas” reporting

Table 2 shows the fitting of two identical lagged dependent variable Poisson models on the count of “Rendición de cuentas” daily news items for two different series (2017 and 2018).

These models are not nested and regress the dependent count variable by two key independent variables from H2 by assessing a trend related to the budget bill’s yearly revision entry to the parliament (days to and days since), as well as a one-order lag of the dependent variable to eliminate the dependent variable error terms’ autocorrelation, and six dummy variables (seven minus the reference category) indicating the day of the week to correct for a day-of-the-week-related seasonality in news reporting (Keele & Kelly, 2006; Wooldridge, 2010).

Table 2 shows that the predicted day-of-the-week seasonality exists and is statistically significant for most days (p> .05) in both years. Whereas the effects of the lagged depen­dent variable behaves as expected (the count of news in the preview day increases the ones in the next), it is only statistically significant for 2018 (p > .001).

Nonetheless, and more importantly to the model, the trends focusing on the day of entry to the parliament are statistically significant (p > .001) in the expected direction: previous to its entry the count of news items increases as the day approaches, and after the event it decreases as time goes on; thus, H2a and H2b are supported.

Both models behave and perform similarly: better than their intercept-only models, with reasonable pseudo R2 values, no serial correlation, corroborated by Durbin’s alternative run on the same OLS’ models as a proxy, as well as the Poisson’s Pearson residual qq­plot and time series line plot (Cameron & Trivedi, 2009). Regarding the Poisson models’ fit, in both cases they do not pass the Deviance goodness-of-fit test (p > .05) and they both pass the Pearson test (but 2018’s far closer to the p > .05 threshold).

Table 3 presents nested models for the 2018 series, one of them including the World Cup period as an additional explanatory variable (testing H1).

 

Besides our core independent variable, there is no relevant variation in any parameters between these nested models, and the controls for serial autocorrelation and seasonality behave similarly to what was descripted for Table 2 (see Pearson residual plotting and Durbin’s alternative test).

However, the main difference between these models resides in the effect of the World Cup time in the daily reporting of “Rendición de cuentas”: supporting H1, it has a statisti­cally significant negative impact.

Moreover, the inclusion of this variable increases not only the explanatory power of the model (being the pseudo R2 comparable as they are nested), but also reduced the Bayesian Information Criterion (BIC) by 2.444 points and made it pass both goodness-of-fit tests, thus providing significant evidence of the improvement of the model’s fit as a whole (Cameron & Trivedi, 2009).

Discussion and Conclusion

The present article aimed to provide empirical evidence of an agenda setting scenario in which journalists and/or news media’s editors could have shifted their agenda by “anticipating or estimating public interests” (Russel et al., 2014, p. 134) on two competing events.

By using the strengths of a natural scenario in which government transparency issues (budget bill’s yearly revision) coincide by chance and only sporadically with FIFA’s World Cup, this study was able to provide statistically significant evidence of news media outlets reducing the report of the transparency issues due to mega-sport events where fútbol is critical for the national identity of the country. We did so by fitting a lagged dependent variable Poisson model on the count of “Rendición de cuentas’” daily news items of the three main digital news media outlets in Uruguay (El País, El Observador and Montevideo Portal).

These models, it provided statistical evidence for: a) H2 that trends in the amount of ag­gregated “Rendición de cuentas” daily news items were critically determined by the day of the budget bill’s entry to parliament; and b) a significant negative effect of the period in which the World Cup was occurring in the daily count of daily news items of “Rendición de cuentas” compared to the previous year. In other words, there is statistical evidence that during the World Cup, news media sites reduced their reporting of the budget bill’s entry to parliament compared to previous years.

The article, thus, presents strong evidence of a lesser-studied agenda setting phenom­enon with relevant consequences for how social and political issues are covered by news media. Without clear potential conflicts of interest that could explain the shift in the reporting, our findings signal the necessity to consider the externalities of conducting critical democratic debates during mega sporting events.

Following a considerably old public policy concern, presented 48 years ago by Cobb and Elder (1971), if the politics of agenda building are to be considered, the timing of the political agenda’s objects also need to be considered. Whereas we argue for the need for news media outlets to monitor and/or prepare in advance to counter the influence of big sports events over other socially critical agenda topics, this is not a media-only issue. Not only do journalists and newsrooms operate under complex and stressful conditions that may reduce the chances of these precaution taken place, but also government officials and civic servants are accountable for these opposing agenda objects issues. Perhaps the measures taken to avoid this type of phenomenon could arise from the parliament or government itself, more so in countries such as Uruguay in which “fútbol” plays a key role in national identity.

Finally, the work has several limitations. First, this study only assessed one of the more external layers of the agenda setting phenomenon, and only from the news media side of the equation. Future studies within this research line should include social media data concerning the issues (see Russel et al., 2014) or directly inquiring individuals’ prefer­ences (see Mitchelstein et al., 2017); nonetheless this is contentious. Among some of the issues discussed by Aruguete (2017), the author provides evidence both in favor and against the idea that the Internet and social media can be considered a reflection of public opinion. Moreover, a vast corpus of digital inequalities literature clearly signals that as biases and socioeconomic inequalities shape digital engagement, it is not adequate to assume that websites and social media reflect public opinion (see Dodel, 2015; Robinson et al., 2015). More so, the issue is considerably problematic because even popular political surveys with big samples, such as online opt-in panels, present serious biases that hinder adequate representativeness and quality of these studies (Hargittai & Karaoglu, 2018). Future inquiries could delve into more complex levels of agenda setting issues, such as framing or networks (McCombs & Valenzuela, 2007; McCombs et al., 2014). Moreover, this study looked into how two agenda-setting events are affected, but more research should be conducted on the relation between broader agenda themes.

Second, this study opted to fit simple and parsimonious models, which need to be im­proved in future analyses. The researchers discarded the whole time series to focus on a closer time frame. The role of other political events such as corruption scandals, mid­term elections and other policy issues, should be assessed in their impact on the budget bill’s yearly revision reporting.

Third, as Keele and Kelly (2006) argue, there is more to dependent variable lagged models than just the violation of an estimator assumption, lagged variables could be also understood as “…a way to view autocorrelation as a potential sign of improper theoreti­cal specification”. The scope of our work did not comprehensively address any news cycle-related hypothesis. Further work should assess and correct for these potential issues.

Nevertheless, despite these limitations, this article sheds light on the need to study and address the caveats of social and political issues coverage by news media in the wake of mega sporting events.

NOTES

1. Lagged dependent variable models are linear or non-linear models with corrections for autocorrela­tions or, in terms of Keele & Kelly (2006), a “as a special case of a more general dynamic regression model that is designed to capture a particular type of dynamics in the dependent variable.” As this study’s dependent variable refers to a count of a phenomenon, a Poisson model was fitted; thus, the term Lagged dependent variable Poisson model.

References

Amoedo, P., & Queirolo, R. (2013). AmericasBarometer Insights: 2013.

Bovens, M. (2007). Analysing and assessing accountability: A conceptual framework. European Law Journal 13(4): 447–468.

Aruguete, N. (2017). Las hipótesis de la agenda setting en el nuevo entorno mediático. Comunicación y sociedad, (28), 35–58.

Cameron, A. C., & Trivedi, P. K. (2009). Microeconometrics using stata. College Station, TX: Stata Press.

Cobb, R. W., & Elder, C. D. (1971). The politics of agenda-building: An alternative per­spective for modern democratic theory. The Journal of Politics, 33(4), 892–915

Evans, M. D., & Kelley, J. (2002). National pride in the developed world: Survey data from 24 nations. International Journal of Public Opinion Research, 14(3), 303–338.

Dader, J. L. (1990). La canalización o fijación de la “agenda” por los medios. In A. Muñoz Alonso (Ed.), Opinión pública y comunicación política (pp. 294–318). Madrid, Spain: Eudema.

Dodel, M. (2015). An analytical framework to incorporate ICT as an independent vari­able. In A. Chib, J. May & R. Barrantes (Eds.), Impact of information society research in the global south (pp. 125–144). Singapore: Springer.

Gift, T., & Miner, A. (2017). “Dropping the ball”: The understudied nexus of sports and politics. World Affairs 180(1), 127–161.

Hargittai, E., & Karaoglu, G. (2018). Biases of online political polls: Who par­ticipates? Socius: Sociological Research for a Dynamic World 4, 1–7. https://doi. org/10.1177/2378023118791080

IAB Uruguay. (2016). Ranking IAB UY Global – Enero 2016 – Redes y propiedades. Retrieved from http://www.iab.com.uy/media/uploaded_files/medicion/IAB_Uruguay_En­ero_2016.xls

Jacobs, S., & Schillemans, T. (2016). Media and public accountability: Typology and exploration. Policy & Politics, 44(1), 23–40.

Keele, L., & Kelly, N. J. (2006). Dynamic models for dynamic theories: The ins and outs of lagged dependent variables. Political Analysis, 14(2), 186–205.

Kim, S. H., Scheufele, D. A., & Shanahan, J. (2002). Think about it this way: Attribute agenda-setting function of the press and the public’s evaluation of a local issue. Journal­ism & Mass Communication Quarterly, 79(1), 7–25.

Lawrence, T. B., and Suddaby, R. (2006). Institutions and Institutional Work. In S. Clegg, Hardy, T. Lawrence & W. Nord, (Eds.). Handbook of Organization Studies (2nd ed., pp. 215–254). London, UK: Sage.

Maneiro, C., & Marchi, W. (2015). Heroes and villains in Uruguayan soccer (2010-2014): A discursive approach. Physical Culture and Sport. Studies and Research, 66(1), 28–42.

McCombs, M. (2005). A look at agenda-setting: Past, present and future. Journalism Studies, 6(4), 543–557.

McCombs, M. E., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36(2), 176–187.

McCombs, M., Shaw, D. L., & Weaver, D. (1997). Communication and democracy. Exploring the intellectual frontiers in agenda-setting theory. Mahwah, NJ: Lawrence Earlbaum Associates.

McCombs , M. E; Shaw, D. L, & Weaver, D. H. (2014). New directions in agenda-setting theory and research, Mass Communication and Society, 17(6), 781–802. 10.1080/15205436.2014.964871

McCombs, M., & Valenzuela, S. (2007). The agenda-setting theory/La teoría agenda-setting. Cuadernos.Info, (20), 44–51.

Mitchelstein, E., Boczkowski, P. J., Andelsman, V., & Leiva, S. (2017). Política y pelotas: Los intereses temáticos de los usuarios de sitios de noticias líderes en Argentina en 2016. CIC. Cuadernos de Información y Comunicación, 22.

Djerf-Pierre, M., Ekström, M., Håkansson, N., & Johansson, B. (2014). The mediatization of political accountability. Journalism Studies, 15(3), 321–338. 10.1080/1461670X.2014.889473

Moraes, J. A., Chasquetti, D., & Bergara, M. (2005). The political economy of the budget­ary process in Uruguay. Inter-American Development Bank. Retrieved from http://www. decon.edu.uy/publica/2005/Doc1905.pdf

Robinson, L., Cotton, S. R., Ono, H., Quan-Haase, A, Mesch, G. S., Chen, W.,  Schulz, J., Hale, T. M., & Stern, M. J. (2015). Digital inequalities and why they matter. Information, Communication & Society, 18(5), 569–582.

Rogers, E., & Dearing, J. (1988). Agenda-setting research: Where has it been, where is it going? Annals of the International Communication Association, 11(1), 555–594.

Russell Neuman, W., Guggenheim, L., Mo Jang, S., & Bae, S. Y. (2014). The dynamics of public attention: Agenda-setting theory meets big data. Journal of Communication, 64(2), 193–214.

Van Hilvoorde, I., Elling, A., & Stokvis, R. (2010). How to influence national pride? The Olympic medal index as a unifying narrative. International Review for the Sociology of Sport, 45(1), 87–102.

Strömberg, D. (2015). Media coverage and political accountability: Theory and evidence. In S. Anderson, J. Waldfogel & D. Strömberg Handbook of Media Economics (vol. 1, pp. 595–622). Amsterdam, Netherlands: Elsevier.

Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. Cambridge, MA: MIT Press.

Matías Dodel is a Sociologist (PhD, M.A. & B.A.) and Psychologist (B.A.) who special­izes in Information Society, Public Policy and Inequality issues. Head of the Secretary of Research at Universidad Católica del Uruguay (UCU), he is also an Adjunct Professor at the Department of Communications. In 2013 he was appointed as the director of the Research Group on Uruguay, Society & Internet (UCU) were he leads the World Internet Project, DiSTO and Kids Online`s Uruguayan Chapters. He also works as a consultant at the National E-government Agency (AGESIC) of the Uruguayan Presidency, were he developed the official ICT surveys.

Federico Comesaña is a Uruguayan Journalist and Entrepreneur specialized in Eco­nomics and Data Science. He is Managing Director and Co-Founder of Enia (Artificial Intelligence and Data Science) and columnist for the newspaper El Observador (Uru­guay), where he was Senior Editor of Economics and Finance (from 2014 to 2017). He is also columnist in TV and radio. He teaches and is a researcher for the Communication Department of Universidad Católica del Uruguay. He studied Economics at Universidad de la República, has a Diploma in Political Economy of Contemporary Uruguay from  CED and a specialization in Data Science at Coursera/Johns Hopkins University.

Daniel Blanc is a Sociologist (B.A.) from the Universidad Católica del Uruguay (UCU), specialized in media, labor market and social security. He is a technical assistant for the National Public Education Administration (ANEP) of Uruguay and a technology consul­tant for private companies such as Mercoplus, where he also works as a data analyst. He has done research on transparency, media, and the future of work and its impact on the Uruguayan social security system. He is a volunteer in the civil association CIVICO, which promotes actions and policies in favor of greater governmental transparency and accountability.