Introduction

Donald Trump’s victory in the 2016 elections in the United States revived a discussion about the role of the press in a democracy and approaching candidates during campaigns (^{Tabor & Wise, 2016}), which became evident again during his re-election bid (^{Pilkington & Gabbatt, 2020}). This discussion was transferred to the televised presidential debates and the role of the moderator, because of the first debate between Biden and Trump, due to the level of aggressiveness that it reached since the sitting president did not respect the rules of the encounter (^{Patten & Johnson, 2020}).

The form that the coverage of the candidates takes in a particular campaign and how they are approached in a live television space have been permanent challenges in several countries. For example, ^{Marien et al. (2020}) show that far-right populist speeches have decreased the deliberative quality of televised debates in Europe, while ^{Montez and Brubaker (2019}) report an increase in aggressiveness in the discourse of US candidates between 2015 and 2016.

The literature identifies political coverage trends that threaten its quality with harmful consequences for democracy (^{de Vreese et al., 2017}). One of them would be covered as a strategic game, which increases the public’s cynicism, contributes to the decline of knowledge about politics, generates a negative evaluation and mistrust (^{Zoizner, 2021}). The journalistic coverage after the debates tends to contain this logic of games, which has repercussions in a lesser understanding from the citizens about the topics discussed (^{Pingree et al., 2012}).

Added to this is the focus on the personal characteristics of the candidates (personalization), which also influences the televised debate (^{Zamora Medina & Rebolledo, 2019}), and the more significant space given to soft news, which would leave the public with little substantial information during the campaign (^{Reinemann et al., 2012}). Although there is no conclusive evidence that these trends are growing in all Western democracies, it is clear that their presence becomes more constant, noticeable, and visible in the proximity of electoral processes (^{de Vreese et al., 2017}).

Theoretical framework

Methodology

This study begins with the secondary use of a database on press coverage of three Chilean presidential campaigns (1989, 1999 and 2009) (^{León-Porath et al., 2015}). This is a content analysis of the information covered by the respective electoral campaigns. To update this database, another two weeks were recorded from the newspapers selected for this study, for the 2017 campaign, which covered the entire official campaign. At the time of writing this work, that year was the most recent presidential campaign carried out in the country. For the present study, only the information generated by the media itself (e.g., analysis or evaluations made by journalists or that are not attributable to sources) was recorded, discarding the one coming directly from political actors (press conferences, statements to the media, or campaign events description). In total, 1 811 units were analyzed for the four elections. For each one of them, up to three topics could be determined.

In the case of the presidential debates organized by ANATEL, a new content analysis was carried out in which all the journalists’ interventions during the debates of the first electoral round, for each analyzed presidential campaign, were recorded (for 1989 the only debate held, organized by Channel 13, was recorded). To classify them, this study is based on the same codebook of the original project, even the same set of codifiers. In total, 398 interventions by journalists were recorded. The code was organized around nine major topics, which were further subdivided into various subtopics. For our purposes, these codes were regrouped as shown in Table 1. We will understand by agenda the topics selection and the focus given to specific issues by the media.

A reliability test was performed for the “topics” variable as a multiple response group, recoded into seven categories, in a subsample of 530 cases from the entire general database, obtaining a Holsti index of 0.867 and a Scott Pi of 0.813. All the following analyses are carried out by weighing the unit of analysis by its respective length (number of paragraphs in the case of newspapers and number of words in the case of journalists’ interventions) to account for the issuer’s emphasis in each section of their intervention.

Results

The topics used by the analyzed press to cover the 2017 presidential campaign (Table 2) break in three senses with what could have been a trend, according to our previous observations (^{León-Porath et al., 2015}). First, after a clear downward trend, there is an increased focus on policies and programs (from 6.3% in 2009 to 15.2% in 2017).

Second, there is a slight decrease in the focus on strategies, from 31.4% in 2009 to 26.9% in 2017. In the analyzed media, we can see a reduction in the emphasis on the candidates’ personal attributes or private life, which had increased from 2.6% in 1989 to 6.5% in 2009 but returned to 2.6% in 2017. If the three dimensions of personalization proposed are considered together (candidate activity, political attributes, and personal attributes), these fall below the range of 20% to 13.8% in 2017.

There is also an increase in the media’s interest in political issues, among which are controversial issues during the campaign and matters of greater political relevance. These had remained in the range of 25% in the three previous campaigns and reached 36.4% in 2017. This is important because this topic clearly displaces interest in campaign strategies to second place for the first time. This situation confirms what was found in the literature on strategic coverage and personalization since it has not been possible to determine a sustained trend towards a uniform increase in these coverages throughout the world. Still, they always remain at a high level.

Regarding the agenda of topics addressed by the journalists in their questions in the presidential debates (Table 2), it can be detected that there is a change in the order of priorities over time. While between 1989 and 1999, the first place of interest was on political topics in general, in 2009, it was displaced to second place due to the increase in interest in public policies and program proposals. A substantial difference can be already established, with the agenda of the analyzed media more focused on campaign strategies. In the debates, there is also less interest in aspects of personalization: the sum of the three indicators in this regard remained above 20% in press coverage until 2009, while in the analyzed debates, it never reached that range.

When analyzing in detail the similarities and differences in the agendas of the coverage of the media with that of the journalists in the debates for the four campaigns for which we have the data (Table 2), we see that the differences, in terms of the percentages assigned to each topic considered, have been significant: they are around a total of 90 absolute points on three occasions and 79 points in 2009. The main difference is in the importance given in the debates to public policy proposals, more neglected in press coverage. The difference started in 1989 with 17.6 points, which increased successively until reaching 41.6 in 2017, despite the increase shown about this issue in the press that year, accumulating a total of 116 points away in the four campaigns.

The second element of contrast is marked by the greater interest of the press in campaign strategies, differences that accumulate a total of 64 absolute points in the four campaigns analyzed. The most significant distance was observed in 2017 (21.2%), although the press reduced its focus on that point. Thus, a situation is configured in which, even though, in principle, in 2017 the press tended to approach the priorities that journalists had shown in the debates of 2009 (increasing the space devoted to public policies and reducing the space dedicated to campaign strategies) the absolute distance increased again. This was due to the concentration of questions in the debate on the candidates’ public policy plans, reaching 56.8% of the total debate topics, a concentration never seen before.

This situation can be explained by a reaction of the journalists participating in the presidential debate, in the face of the scandal produced in public opinion by the primary debate of the center-right coalition in that same campaign, which was marked by crossed recriminations and personal attacks. This would have led the debate’s journalists to focus on topics that they considered most advantageous to the public.

The third aspect that accumulates differences is the treatment of political topics in general, an element on which the journalists in the debates placed great emphasis in 1989 (46.2%) and 1999 (42.6%), not so much the press (around 24%). However, in 2009 both agendas got close when the absolute difference reached only 5.6 points, but they distanced again in 2017 (13 points of difference). This time is due to the decreased focus on these issues during the debates (a situation already explained here) and the increased interest of the press in it.

It also deserves comment on what happened with the personalization. Suppose the three factors into which the variable above was divided for this study (activities, political attributes, and personal attributes of the candidates) are considered together. In that case, it can be seen that in 1989 its total was similar in both agendas (around 10%). In 1999 a particular situation occurred: since the rules forced journalists to ask only common questions, i.e., the same question had to be answered in turns by the two candidates present, it was impossible to ask questions regarding individual aspects of each candidate and/or candidacy. That year, the analyzed media also allocated 14.3% of their topics to the personalization of the campaign, surpassing public policies (10.7%). In 2009, the values were again similar (12% in the media against 9.8% in the debate), and in 2017 there was -for the first time- a slight increase regarding interest in the debates’ topics (12.5%) while in the press fell to 8.2%.

However, if the internal composition among the three indicators is considered, there is a significant difference. For the journalists hosting the debates, the focus on personalization has always been on the political attributes of the candidates. This topic is expected to be of interest to the public since it precisely points to one of its objectives: to test the aptitudes of the presidential candidates for office. Instead, the press, for example, in 2009 focused on the so-called aspects of privatization. Although if the obvious interest of the press in covering the activities of the candidates is deducted, typical of their work of informing, in the other three years analyzed, the press also preferred to highlight the political attributes over the personal ones.

Final discussion

Our study has proven that electoral debates are a particular moment in the coverage of electoral campaigns, at least in terms of the topics journalists propose to the candidates for discussion. In the analyzed case, the agenda that the journalists use to formulate their questions differs substantially from the agenda with which the traditional media, the so-called “serious press”, covers the campaign; devoting more space to the discussion of the “substantial issues” of an electoral campaign, such as public policy proposals, and less to the topics that prevail today in the world’s coverage of campaigns, such as strategic game and personalization.

To validate this test, we must remember that Chile was described here as an extreme case in terms of the leading role journalists have in developing the debate. Due to its adoption of the American experience with the format of the journalists’ panel and because the hosts are faces or anchors of their TV channels, it could be expected that the agendas would be similar to those of the press or that the debate would even be more marked by strategy and personalization issues. In our bibliographical discussion, we have established that the press agenda should not differ much from TV. Still, we have also shown that, in the case of Chile, the TV should be even more focused on these issues, given its heavy commercialization.

Suppose in these conditions we have shown that the agenda of the debates focus more on the substantial issues. In that case, we should expect that countries that do not use the panel of journalists and/or their hosts are not TV personalities; their agendas would also tend to focus more on substantial issues than the general media coverage of the electoral campaign. This is something that now must be empirically demonstrated. Therefore, we can affirm that the debate responds to its own media logic, which is effectively isolated from the daily journalistic routine, becoming an informative and deliberative space on the essential issues of the campaign, thus fulfilling its character as a civic ritual, probably because in Chile the journalists present at the televised debates have tended to develop a journalistic style of a civic nature (^{Núñez-Mussa, 2018}) focused more on knowing what the plans of the presidential candidates are, rather than about topics that could be considered controversial such as ideals, values, or personal qualities.

This is consistent with ^{Schroeder’s proposal (2000)}, which maintains that the debates, by not depending on the logic of the commercialization of the media, can generate a difference in the way of covering the candidates and their issues; thus, preserving its democratic character and maintaining the objective of being a communicative instance where potential voters, especially undecided ones, can learn about and contrast the proposals of the candidates in a limited space and time. In any case, it is necessary to keep a permanent discussion about the formats and regulations since the tradition and existence of the event itself does not guarantee that it fulfills that function, as demonstrated by the debates of the US election in 2020.

As a final reflection, it is possible to say that, despite the growing imposition of commercial logic in the media, the presidential debates in Chile continue to be a space that favors questions that deal with public policies over issues related to character. Probably because these debates, being broadcast on all television channels, reduce competition between stations and ensure a high rating. This accounts for the value of the instance within the democratic process.

At the same time, our work complements a previous one on the media agenda in electoral campaigns (^{León-Porath et al., 2015}). It confirms that, in the case of Chile, the pre-eminence of issues such as strategic coverage or personalization, despite having an important impact on the press agenda, is not constantly growing. It presents setbacks, which accounts for a moving nature. This reaffirms the need to analyze which variables explain these changes in the interest of journalistic coverage compared to political or substantial issues of the campaign, which requires closer observation of specific cases.