Issue convergence and issue divergence in the content of television campaign messages in two-round presidential systems: Chile, 2005–2017

Issue convergence theories posit that, in a two-person race, the messages of the candidates converge to the same priorities. Conversely, issue ownership theory argues that candidates will emphasize issues where they, or their parties, hold a reputational edge. In multi-candidate, two-round presidential elections, the first-round campaign normally allows for greater issue divergence. However, in the runoff campaigns, the two candidates should be more likely to converge on the issues. Through content analysis of television ads, this study examines four Chilean presidential election campaigns from 2005 to 2017. Employing a regression discontinuity design (RDD), we explore if candidates converged on the issues mentioned in their campaign messages in the runoff compared to the first round. The findings indicate that right-wing candidates more frequently address crime, pensions, and employment, whereas left-wing candidates focus more on education during the first-round campaign. In the runoff, we find convergence on campaign messages on education and pensions, but in other issues, the candidates’ messages continue to show divergence.

Introduction

Issue convergence theory suggests that in a two-person race, campaign messages will align with the electorate’s top priorities. Conversely, issue ownership theory predicts that candidates will highlight issues where they or their parties are deemed more trustworthy. Based on the long standing literature that discusses the differences in campaign strategies and dynamics in democracies with two-round presidential elections, we expect that the campaign messages of presidential candidates in the first round will present more issue divergence and prioritize issue ownership while the campaign messages in the ballotage will be more likely, following the need to cater to the median voter, to present issue convergence.

This study examines the case of Chile, a multiparty presidential system with runoff provisions, where candidates receive equal free daily television time during the campaign. We explore the mentions of issues in the television campaigns by the candidates that made it to the runoff in every election round to identify whether there is convergence on some issues in the runoff campaign. In 2017, right-wing candidate Sebastián Piñera averaged 0.53 daily mentions of crime in the first-round campaign. In the 12 days of the runoff campaign, Piñera’s television ads mentions of crime increased to 1.47 times per day. In contrast, his center-left opponent, Alejandro Guillier, mentioned crime 0.18 times per day in the first round and 0.07 in the runoff. With crime being a significant concern for voters that year, the two candidates diverged in their mentions of crime in their television ads. Piñera leveraged the right-wing’s perceived issue ownership advantage on crime and secured victory in the runoff with 54.6% of the vote.

Across four presidential campaigns from 2005 to 2017, we analyzed 29.7 h of campaign ads aired during the 40-day franja electoral in Chile. We compare the daily average mentions of different issues in television ads by the rightwing and leftwing candidates in every election cycle to identify convergence and divergence on six different issues in the first-round campaign and in the runoff campaign. Using a regression discontinuity design (RDD), we also explore if the mentions of six different issues between the first-round and runoff in the television campaign ads.

In the ensuing sections, we delve into the concepts of issue convergence and divergence in campaign messages. We postulate a hypothesis and proceed to test it by comparing the mentions candidates made of six issues in their first-round and second-round television ads. We report issue convergence in the runoff campaign television ads, but only on certain issues. There is evidence of issue divergence in both electoral rounds, as candidates strategically emphasize issues where their parties possess issue ownership.

The impact of television ads on presidential campaigns

Defined as “the process of planning and executing activities in an attempt to win votes” (Simon, 2002, p. 35), campaigns are central to elections, although the technological tools available have changed over time and the means of communication between candidates and voters constantly evolve (Graber, 2001). Since they were first incorporated into the repertoire of electoral campaigns in the 1960s, television ads have remained an important component of presidential campaigns (Petrocik et al., 2003). Though in recent years, online networks and phone apps have become an increasingly popular tool in campaigns (Borah, 2016; Aparaschivei, 2011), television remains an important media for candidates to communicate with voters, especially in lesser developed countries (Boulianne, 2015).

In Latin American emerging democracies, television ads have also played a critical role in presidential campaigns (Boas and Gans-Morse, 2009; Boas, 2005; Boas, 2010; De la Torre and Conaghan, 2009). Although the impact of television ads on vote choice is a subject of debate (Kaid and Johnston, 1991; Franz et al., 2020), the importance of television ads in electoral campaigns cannot be understated, especially in emerging democracies where there is more limited access to online networks and where candidates have free access to prime time television to disseminate their message (Boas, 2010, 2016; Posner, 2003).

The study of the interaction between television ads and electoral campaigns has focused on the impact of the former on electoral results (Boas, 2010; Boas, 2005; Sides et al., 2022; Fowler et al., 2016). Here we focus on changes in issues prioritized in the content of television campaign messages between the first-round and runoff campaigns in presidential elections.

Issue divergence and issue convergence in campaign messages

The issue ownership theory postulates that, because of the policies they have pursued in the past, political parties have a stronger reputation as being more competent and credible in handling some issues (Petrocik, 1996; Budge and Farlie, 1983; Petrocik et al., 2003). As Damore explains, “the theory of issue ownership rests on the proposition that because of long standing party reputation, candidates’ are more likely to be perceived as credible over issues owned by their party […] For instance, voters see Republicans as being better able to handle foreign policy and government management issues, while the Democrats are perceived as more capable on civil rights and social welfare issues” (Damore, 2004, p. 392). Some studies suggest that issue ownership is based on the policy priorities of party elites more than on the experiences of voters with how parties have handled certain issues when in control of the government (Fagan, 2021). Issue ownership advantage leads candidates to produce messages that resonate with the public based on past experiences and expectations. As Damore argues, “candidates should focus their attention on issues that either play to their strengths or their opponents’ weaknesses” (Damore, 2004, pp. 391–392). Issue ownership might be extended to individual candidates. Where the rules generate incentives to cultivate a personal vote (Cain et al., 2013), individual traits and experience can be useful for candidates to claim issue ownership over certain priorities (Greene and Baker, 2011).

Issue ownership can be associated with two dimensions that are analytically different: competence, when the party has shown to be better at handling an issue, and association, when people have observed a party championing an issue over time (Walgrave et al., 2012; Walgrave et al., 2015). The sources of issue ownership are associated with partisanship, attitudes, perceptions of the societal development as indicating parties’ performance, and perceptions of linkages between specific parties and specific constituency groups (Stubager and Slothuus, 2013). Studies show that issue ownership changes over time and that parties can impact the issues they own through strategic campaigning (Walgrave and Lefevere, 2017) and resist engaging with issues raised by their opponents where the opponents presumably have electoral advantages. More recently, some have questioned traditional definitions of the concept by showing that issue ownership is comprised of multiple distinct dimensions of public opinion and cannot be reduced to a single concept independent of other political concerns (Therriault, 2015). But in general, when asked about issue positions, voters easily associate specific issues with specific parties (Goggin et al., 2020).

Candidates aspire to win by priming issues where they exercise ownership and by minimizing issues where their rivals exercise ownership (Sides, 2006; Damore, 2004; Petrocik et al., 2003). Moreover, as public priorities sometimes differ from the issues owned by their parties, candidates will seek to respond to popular priorities, thus resulting in demand-driven issue convergence (Damore, 2005). As the number of candidates in the race affects the dynamics of the campaign, the incentives to exert issue ownership also vary. In multi-candidate first-round elections, candidates will strategically campaign to optimize their chances of making it to the runoff (Cox, 1999). Thus, we would expect candidates in a multi-person race to diverge in their campaign messages in the first round more than in the runoff election, where there are only two candidates.

Because candidates seek to minimize their rival’s advantage, they might engage on issue trespassing, addressing policy priorities that have been traditionally owned by their opponents (Damore, 2004; Karlsen and Aardal, 2016; Greene and Baker, 2011). If a leftwing party exercises ownership over educational policy or poverty reduction, the rightwing party candidate will need to engage in issue trespassing to respond to popular concerns over those issues (Nickelsburg and Nortpoth, 2000).

The literature on candidates’ campaign strategies has been heavily influenced by the median voter theorem which suggests that, given fixed policy preferences in the electorate, in a two-person race, candidates’ messages converge to the median voter (Downs, 1957; Grofman, 2004; Fowler et al., 2021, p. 132). In a static approach to campaign messaging, candidates seek the support of the median voter to win the election (Simon, 2002, pp. 45–51). To do so, candidates embrace the preferred policy positions of the median voter. Even in democracies where the electorate is polarized, the decisive voters—those that are equated with the median voter—have policy preferences that candidates seek to embrace to win their vote.

However, in a multi-candidate first-round race, convergence to the median voter might not happen or, alternatively, the candidate favored by the median voter might not win (Romer and Rosenthal, 1979, p. 146). Multicandidate elections are full of uncertainty resulting from the different candidate strategies (Simon, 2002, pp. 51–55). Varying voter preferences over certain issues, unanticipated events, including scandals, and technological developments that afford candidates new tools to reach voters (Fowler et al., 2021, p. 130) impact the dynamics of presidential campaigns across elections and over time. As a result, presumptive losers in a two-person race might end up winning if they pivot their campaign issues in strategic and electorally relevant ways (Amorós and Puy, 2004).

The number of candidates affects the incentives to cater to the preferences of the median voter in the first-round vote. In turn, in a runoff election, where there are only two candidates, the incentives to cater to the median voter are stronger (Cox, 1999). Thus, while the literature on issue ownership centers on the factors that lead candidates to differentiate on their campaign messages, explanations on what leads candidates to embrace campaign messages that are based on the median voter theorem point to the motives behind the candidates adopting policy positions and campaign pledges that mirror those of the decisive voters in the election. Thus, while the literature on issue ownership admits that candidates occasionally engage in issue trespassing, the literature that emanates from the median voter theory expects that the electoral incentives to convince the median voter will inevitably produce centripetal effects in the messages that candidates prioritize in their campaigns. For the lack of a better concept, we refer to those centripetal incentives of the median voter theory as issue convergence. This brings us to our hypothesis:

Hypothesis: There is more issue divergence in the first round and more issue convergence in the runoff campaigns.

The case of Chile, 2005–2017

Prior studies on presidential campaigns in Chile provide a descriptive view of the issues prioritized by the candidates (López-Hermida Russo, 2017) or highlight idiosyncratic issues, like gender affinity (Morales, 2008; López-Hermida, 2009), newspaper coverage of candidates (Porath et al., 2015; Portales, 2009), and the online reach of campaigns (Cárdenas et al., 2017). A study on the impact of free television time reports high citizen approval for that campaign feature and declining, but still significant, impact for candidates in reaching national audiences (Uribe et al., 2018). Another study describes the presence of issue ownership in television ads in presidential campaigns (Lisbona and Navia, 2018). Television has been widely used in Chile as a mechanism to access political news (Valenzuela and Arriagada, 2011)—even given the growing of usage of online networks (Espinoza Bianchini et al., 2021). The consumption of media impacts vote choice, especially among the youth (Scherman and Arriagada, 2012).

Since 1980, Chile has had a multiparty presidential democracy. Presidential elections use majority rule, with runoff provisions if no candidate receives more than 50% in the first-round vote. Thus, electoral rules are conducive to the formation of large multiparty coalitions—at least in the presidential election runoff. When they complete a four-year term, presidents can run again after sitting out for one term. The leftwing Concertación coalition won the presidency 4 consecutive times between 1989 and 2005 and again in 2013 (under a new name, New Majority). In 2009 and 2017, the rightwing coalition, Alianza, won the presidential election. Table 1 shows presidential election results for the 4 elections held between 2005 and 2017. In 2005, there were 4 first-round candidates—two from the Alianza. Concertación’s Michelle Bachelet received the highest vote share in the first round and went to defeat Alianza’s Sebastián Pinera with 53.5% of the vote in the runoff. In 2009, Piñera received a plurality in the first-round vote among 4 candidates and defeated the Concertación candidate, former president Eduardo Frei (1994–2000) in the runoff. In 2013, New Majority’s Bachelet received a plurality in the first round among 9 candidates and defeated Alianza’s Evelyn Matthei in the runoff. In 2017, among 8 candidates, Piñera won a plurality in the first-round vote (36.6%) and defeated New Majority’s Alejandro Guillier in the runoff.

Table 1

Table 1. Presidential election results in Chile, 2005–2017.

Thus, the 8 candidates that made it to the runoff in those 4 elections were from the Concertación/New Majority (NM) and Alianza coalitions. In three of those elections, the Alianza candidate in the runoff was Piñera. In 2005 and 2013, the Concertación/NM candidate was Bachelet. In 2013 and 2017, the winning candidates were former presidents, Bachelet and Piñera, respectively. In 3 of the 4 contests, the results produced alternation in power, as the ruling coalition candidate lost in the runoff with the candidate of the opposition coalition.

Chilean law prohibits the purchase of television time for campaigns, mandating instead a 20-min free prime time slot on all open access networks for 28 days before the first-round vote and 12 days before the runoff (Uribe et al., 2018; García Rodríguez, 2013). This allocation, divided equally among candidates, varies per election based on the number of contenders. Television time gives candidates autonomy over the content of their messaging, in contrast with debates or interviews where the agenda is set externally. Additionally, the lead time for ad production means candidates cannot swiftly alter their messages in reaction to competitors or campaign developments.

Understanding the factors that drive changes in campaign message content and priorities during the frenetic pace of election season is complex. Candidates must navigate a myriad of influences including real-time events, both positive and negative feedback, poll results, and their own perceptions and intuitions. Moreover, the production cycle of television ads introduces a lagged effect, with content often being finalized days before airing.

The ads, sourced from the National Television Council, provide data for 8 candidates. Table 2 details the television time distribution for each candidate across four election cycles, highlighting the inverse relationship between the number of candidates and the individual airtime allotted.

Table 2

Table 2. Minutes of television ads for each presidential candidates in Chile, 2005–2017.

Methodology

We explore whether the mentions of policy priorities in the television ads of the presidential candidates in four election rounds in Chile changed between the first-round and the runoff campaigns. We first identified the policy priorities for Chileans by looking at public opinion polls. The Center for Public Studies (CEP), a private public interest think-tank, has conducted national polls for over 40 years. CEP polls include a question on which 3 issues people believe should be the most important priorities for the government. The closed list of policy priorities includes crime, pensions, health, education, salaries, jobs/employment, poverty, corruption, drugs, housing, inequality, judicial system, environment, transportation, human rights, and constitutional reforms. Over the years, the list of policy priorities has varied. We reviewed CEP polls since 2002 to identify the 5 issues that were constantly among the most important for public opinion in the 4 election cycles. In addition, we included pensions, a policy priority added to the CEP polls before the 2017 election cycle. Since pensions immediately climbed as a top priority for Chilean public opinion, we added it to the 5 other priorities. Since polls were conducted a few months before each election, we assume that those priorities remained in voters’ minds on election day.

Figure 1 shows the policy priorities that were most important over the years (education, health, crime, salaries, jobs) and pensions, the most important priority for the 2017 election cycle. For visual purposes, we grouped the 6 priorities into social and economic priorities. Crime remained the most important priority over the four election cycles. Education also experienced an upward trend, reaching a peak in 2013—when Bachelet ran for a second term. Health also gained relevance over time. On economic issues, jobs became decreasingly relevant. By 2013, salaries had become the leading concern.

Figure 1

Figure 1. Public opinion priorities for government action in Chile, 2002–2017. Source: Authors, with data from Center for Public Studies.

Following Petrocik (1996, p. 832), we classify the 6 public opinion priorities according to whether they are owned by left or rightwing parties. In Chile, rightwing parties tend to promote economic growth and employment as poverty reduction and economic inclusion tools. In turn, leftwing parties favor a state-centered approach and promote government subsidies and social programs. For that reason, we consider Education, Health, and Pensions as issues where leftwing parties exert ownership and Jobs/Employment, Crime, and Salaries as issues where rightwing parties exert ownership.

To determine whether presidential candidates mentioned these priority issues in their campaigns, and building on previous research that coded issue mentions in television ads by party association (Petrocik et al., 2003; Lisbona and Navia, 2018), we coded the television ads for the 28 days of the first-round campaign and the 12 days of the runoff campaign for the 8 candidates that made it to the runoff in the four presidential elections. Thus, the unit of analysis is the day of free television ads during the presidential campaign for each presidential candidate. In total, we have 320 daily observations. We estimate a different model for each of the six public opinion policy priorities.

To code the issues mentioned in the ads, we recruited 10 political science students from our university in May of 2018. We provided guidelines and then allowed coders to work independently. Coders were provided with the entire set of television ads for the 8 candidates and asked to watch the ads independently to code the messages. Altogether, the coders watched 892.2 min of televised campaign ads. Following Boas (2015), coders were instructed to identify if the mention in the television ad was on public policy, the candidate’s personal traits, or political divides present in Chilean society. Policy issues are those associated with legislative initiatives, executive action, or government goals. Coders were asked to pay attention only to the spoken and written messages in the ads—not to images, symbols, or other forms of communication. We sought to replicate a real-life scenario where different viewers interpret television ads differently. We sought to replicate what television viewers would have interpreted as references to specific campaign issues in the ads, not what would be accurately reported by objective coders.

We averaged all the mentions recorded by the coders for each issue each day. While our approach is prone to an overreporting of mentions by some coders that identify many mentions of any given issues, alternative coding techniques—such as counting only the references that at least 80% of coders report—might risk underreporting mentions of specific issues. Thus, if the 10 coders reported one mention of education on a given day, the value for that candidate that day is 1. Since candidates can mention the same issue more than once, the range goes from a low of 0 to a high of 5.2 mentions per day. While some coders can interpret that a candidate made one or more references to crime in the television ads, others might not see those mentions. Table 3 shows the number of mentions and the daily average for the 8 candidates in the 4 presidential cycles. To complement this information, Table A1 in the appendix shows the average mentions for the first-round and runoff campaigns for each of the 8 candidates indicating whether the candidates were left or rightwing, belonged to the incumbent coalition or were opposition candidates, and whether they were leading or trailing in the polls.

Table 3

Table 3. Mentions of public policy priorities in television campaign ads in Chile, 2005–2017.

Our dependent variable is the average number of daily mentions of each priority issue. We compare the average number of daily mentions of each of the 6 issues in the first-round for the 4 leftwing and 4 rightwing candidates with the number of daily mentions of the 6 issues in the runoff. We expect that the number of mentions will converge in the runoff between leftwing and rightwing candidates.

Additionally, we estimate a regression discontinuity design (RDD) to explore whether the focus on certain issues changed between the first-round and the runoff campaigns. Formally, a RDD allows us to identify the effect of the treatment assuming that all the other variables are randomly assigned on the closest area to the cutoff or treatment point (Shadish et al., 2002, pp. 208–211). We include third order polynomials to allow the curves to better capture the data variance over time (Broockman, 2009; Gelman and Imbens, 2019; Imbens and Kalyanaraman, 2012; Imbens and Lemieux, 2008; Lee, 2008; Lee and Lemieux, 2010). The size of the coefficient at the cutoff point is the quasi-causal effect of the discontinuity (Angrist and Pischke, 2009).

RDD has been used for studies on differences in voting behavior in two-round elections in France (Pons and Tricaud, 2018) and on the impact of an electoral victory on the subsequent election (Feierherd and Lucardi, 2022). We innovate by exploring whether the focus in the issues prioritized by the candidates changed between the first-round and the runoff campaigns. In our design, we expect a discontinuity between the first-round and runoff campaign priorities. Thus, if the coefficient of the discontinuity indicator in the RDD models are significant, we can safely conclude that the priority of the issue in the runoff campaign ads diverges from the priority candidates gave to the same issue in the first-round campaign ads.

The independent variable of interest is whether the ads were broadcast in the first-round or in the runoff. We expect that candidates will focus on different priorities in their first-round and runoff television ads. We also care about the political ideology of the candidates and thus distinguish between left and rightwing candidates. In the four election cycles, there was a leftwing candidate from the Concertación/NM coalition and a rightwing from the Alianza coalition. Table 4 shows the descriptive statistics for the variables we use in the models and figures.

Table 4

Table 4. Descriptive statistics for dependent and independent variables.

Campaign messages in the first round and runoff in Chile in 2005–2017

Table 5 presents the average daily mentions of social and economic issues in the television campaign messages for the 8 presidential candidates who made it to the runoff in the four election rounds. To be sure, as there are only two candidates in the runoff, there is more television time allotted to each candidate in the runoff than in the first-round campaign. Thus, in general, candidates make more mentions of issues in the runoff than in the first-round campaign. Still, the increase in mentions in the runoff campaign is not uniform across issues. Some candidates choose to prioritize more issues than others relative to the number of mentions they had made in their first-round ads.

Table 5

Table 5. Average mentions of campaign messages by candidate Chile, 2005–2017.

There is convergence in the runoff messages on crime and education. In the runoff, rightwing candidates made more mentions of education, thereby reducing the first-round advantage by leftwing candidates, while the opposite occurred in mentions of crime. Yet, there was no convergence in the runoff on the other four issues. In fact, the gap in mentions of health, jobs, pensions, and salaries was wider in the runoff than in the first-round campaign. As expected, rightwing candidates make more mentions of crime, jobs, and salaries, but contrary to what we expected, leftwing candidates did not make more mentions of health or pensions. Leftwing candidates made more mentions of education, but only in the first-round campaign. In the runoff campaign, rightwing candidates made more mentions of education than leftwing candidates. The data on Table 5 is only partially consistent with the expectation that there will be issue convergence in the runoff campaign. We observe issue convergence, but only on crime and education, not on the other 4 issues.

The results in Table 5 also presents partial evidence that candidates take advantage of issue ownership in their campaign mentions. While rightwing candidates promote the issues where they exercise ownership—like crime, jobs and salaries—leftwing candidates make more mentions of health and education—issues that the literatures associated with leftwing parties—but only in the first-round campaigns.

For a more reader-friendly representation of the results in Table 5, Figure 2 shows the change in the absolute value of the difference in the mentions of campaign issues between the leftwing and rightwing candidate in the first-round and runoff campaigns. There was convergence in the first round and the runoff campaigns in the number of mentions of crime and education, but divergence in mentions of health, jobs, pensions, and salaries. Notice that since there are only two candidates in the runoff—as opposed to several candidates in the first-round vote—one could expect that, given that candidates have more time in the runoff than in the first round, candidates will systematically make more references to all issues in the runoff campaign than in the first-round campaign ads. Yet, as Figure 2 shows, when compared to the mentions in the first-round ads, candidates increased the number of mentions in the runoff in some issues and decreased the number of mentions in the runoff in other issues.

Figure 2

Figure 2. Daily mentions of selected issues in first-round and runoff campaigns for leftwing and rightwing presidential candidates in Chile, 2017–2021. Absolute value of the difference in mentions between leftwing and rightwing candidate. Source: Authors based on data coded from presidential television ads in Chile in each election.

Regression discontinuity design analysis

Table 6 shows RDD models on the mentions in the television ads of the six priority issues in the presidential campaigns in Chile. The results show that the mentions of education and pensions present a discontinuity between the first-round and the runoff ads. Candidates made more mentions of those issues in their runoff campaign ads than in the first-round ads. In turn, the other issues—crime, health, jobs and salaries—present no discontinuity between the first-round and the runoff campaigns. Two of the six RDD models are consistent with the claim that candidates update their campaign priorities between the first round and the runoff, altering some of the emphasis they place on the issues they mention on their television ads.

Table 6

Table 6. RDD models for public opinion priority issues presidential campaign messages in the first and runoff rounds in Chile, 2005–2017.

The results also show that rightwing candidates make more mentions than leftwing candidates of crime, jobs, pensions and salaries. In turn, leftwing candidates make more mentions of education than rightwing candidates. This is somewhat consistent with the claim that rightwing candidates exert issue ownership over economic dimensions and crime while leftwing candidates dominate over social issues. Pensions can be considered a social issue, but in the context of Chile, where the pension system was privatized more than 40 years ago, the debate over pensions crosses traditional boundaries as people save money for their retirement in individual accounts. In recent years, the state has acquired a more active role in supplementing individual pensions with a solidarity pillar funded with the public sector budget. Still, with the possible exception of pensions, the models show results that are consistent with the claim that candidates exert issue ownership. Rightwing candidates emphasize economic issues and crime-related issues in their campaigns and leftwing candidates emphasize social issues, like education.

In the appendix, we included two additional models with different cutoff points to identify possible discontinuity. We do so to confirm that the discontinuity identified in Table 6 is not just a random observation. Table A2 presents the results of the first RDD model, with an 8-day cut point for the social and economic issues respectively, and Table A3 presents the results for the RDD model with a 32-day cut point. As expected, neither model shows significant impact on the content of the television campaign messages.

For an easier visualization of the results, we plotted the values for the observations—the average number of mentions of every issue on a single day by all the respective candidates—and estimated two linear models with polynomials to identify the presence of a discontinuity between the first-round and runoff campaign messages. In the figures, we use the same scale for all the issues to simplify the visuals. Figure 3 shows the RDD results for the differences in campaign ads for leftwing candidates and rightwing candidates on the selected social and economic priorities. There is evidence of issue ownership and issue convergence in the runoff campaigns. For example, leftwing candidates made more mentions of education in the first-round campaign. But there is a discontinuity in the runoff campaign whereby all candidates make more mentions of education in the runoff ads than in the first-round ads. As education is identified in the literature as an issue where leftwing candidates exercise ownership and, in the case of Chile during the period education was a priority for voters, as reported in the CEP polls, the evidence from Figure 3 shows that candidates in the runoff converged in their message as they equalized their mentions of education in their campaigns. Thus, on education, the evidence is consistent with our hypothesis. Leftist candidates make more mentions of education, but all candidates also make more mentions of education in the runoff campaign than in the first-round campaign.

Figure 3

Figure 3. RDD on campaign ads by the candidate’s ideology, Chile, 2005–2017. Source: Authors based on data coded from presidential television ads in Chile in each election.

On crime, rightwing candidates make more mentions in their first-round campaign than leftwing candidates, but there is no discontinuity between the first-round and runoff campaigns. Thus, the evidence is inconsistent with the issue convergence expectations and consistent with the issue ownership claims. On health, there is no issue ownership difference between left and rightwing candidates and no discontinuity effect between the first-round and runoff campaigns. Health is an important concern for the population, but candidates did not present differences in how they primed the issue in the first-round and runoff campaigns or whether they were left or rightwing candidates. On pensions, there is evidence of issue ownership, but unlike what we expected, rightwing candidates make more mentions of pensions than leftwing candidates—who supposedly exert issue ownership on pensions. There is also evidence of a discontinuity effect between the first-round and runoff campaigns. Candidates make more mentions of pensions in their runoff campaigns than in the first-round messages. On jobs and salaries, there is evidence of rightwing candidate issue ownership, but no discontinuity effect between the first-round and the runoff. Rightwing candidates make more mentions of those issues, but there is no significant change between the first-round and the runoff campaigns in how candidates prime those issues.

Conclusion

We leverage a case of a presidential system with runoff provisions where candidates have access to free television time to broadcast their campaign messages to explore whether candidates converge or diverge in the mentions of issues in their first-round and run-off campaign messages. We analyze data with mentions of 6 priority issues in the television ads of the candidates who made it to the runoff in 4 presidential election cycles in Chile between 2005 and 2017. We find evidence that candidates converge in mentions of crime and education in the runoff campaign while in health, jobs, pensions and salaries, candidates diverge based on their ideological identification and prior issue ownership.

We show evidence that, in two-round presidential systems, is consistent with the claim that candidates modify their campaign messages between the first round and the runoff in some, but not all issues. While there is evidence of issue ownership in the messages that candidates prioritize in their first-round campaigns, there is also evidence that candidates tend to converge on other issues in their messages in the runoff campaigns. Although the number of overall mentions of all issues increase in the runoff campaign compared to the first-round campaign, the increase is not uniform across issues. Candidates increase the number of mentions of some issues more than the number of mentions of other issues.

Future studies should explore other forms that candidates use to prime priority issues in their television campaign messages—such as visuals or other references beyond spoken words. Studies should also focus on the predominance candidates give to specific issues relative to other issues. The mentions of a specific issue as a share of all the mentions of policy issues a candidate makes in the campaign would allow for a diachronic analysis of the most prominent issues for a candidate in the first round compared to the runoff campaign.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

AL: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. FL: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. PN: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. JC: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. We acknowledged funding from the National Fund for Scientific and Technological Development (grant # 1231627) and by the Millennium Nucleus Center for the Study of Politics, Public Opinion and Media in Chile (grant number NCS2024_007).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Gen AI was used in the creation of this manuscript.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpos.2025.1548670/full#supplementary-material

References

Amorós, P., and Puy, M. S. (2004). “A model of political campaign manipulation.” Fundación de Centro de Estudios Andaluces. Documento de trabajo. 20.

Google Scholar

Angrist, J. D., and Pischke, J.-S. (2009). Mostly harmless econometrics: An empiricist's companion. Princeton, NJ: Princeton University Press.

Google Scholar

Aparaschivei, P. A. (2011). The use of new media in electoral campaigns: analysis on the use of blogs, Facebook, twitter and YouTube in the 2009 Romanian presidential campaign. J. Media Res. 4, 39–60.

Google Scholar

Boas, T. C. (2005). Television and neopopulism in Latin America: media effects in Brazil and Peru. Lat. Am. Res. Rev. 40, 27–49. doi: 10.1353/lar.2005.0019

PubMed Abstract | Crossref Full Text | Google Scholar

Boas, T. C. (2010). Varieties of electioneering: success contagion and presidential campaigns in Latin America. World Polit. 62, 636–675. doi: 10.1017/S0043887110000213

Crossref Full Text | Google Scholar

Boas, T. C. (2015). “Voting for Democracy: Campaign Effects in Chile’s Democratic Transition.” Lat. Am. Polit. Soc. 57, 67–90.

PubMed Abstract | Google Scholar

Boas, T. C. (2016). Presidential campaigns in Latin America: Electoral strategies and success contagion. New York: Cambridge University Press.

Google Scholar

Boas, T., and Gans-Morse, J. (2009). Neoliberalism: from new Liberal philosophy to anti-Liberal slogan. Stud. Comp. Int. Dev. 44, 137–161. doi: 10.1007/s12116-009-9040-5

PubMed Abstract | Crossref Full Text | Google Scholar

Borah, P. (2016). Political Facebook use: campaign strategies used in 2008 and 2012 presidential elections. J. Inform. Tech. Polit. 13, 326–338. doi: 10.1080/19331681.2016.1163519

Crossref Full Text | Google Scholar

Boulianne, S. (2015). Social media use and participation: a meta-analysis of current research. Inf. Commun. Soc. 18, 524–538. doi: 10.1080/1369118X.2015.1008542

Crossref Full Text | Google Scholar

Broockman, D. E. (2009). Do congressional candidates have reverse coattails? Evidence from a regression discontinuity design. Polit. Anal. 17, 418–434. doi: 10.1093/pan/mpp013

Crossref Full Text | Google Scholar

Budge, I., and Farlie, D. (1983). Explaining and predicting elections: Issue effects and party strategies in twenty-three democracies. New York: Taylor & Francis.

Google Scholar

Cain, B., Ferejohn, J., and Fiorina, M. (2013). The personal vote. Cambridge, MA: Harvard University Press.

Google Scholar

Cárdenas, A., Ballesteros, C., and Jara, R. (2017). Redes sociales y campañas electorales en Iberoamérica. Un análisis comparativo de los casos de España, México y Chile. Cuadernos. Info 41, 19–40. doi: 10.7764/cdi.41.1259

Crossref Full Text | Google Scholar

Cox, G. W. (1999). Electoral rules and electoral coordination. Annu. Rev. Polit. Sci. 2, 145–161. doi: 10.1146/annurev.polisci.2.1.145

Crossref Full Text | Google Scholar

Damore, D. F. (2004). The dynamics of issue ownership in presidential campaigns. Polit. Res. Q. 57, 391–397. doi: 10.1177/106591290405700304

Crossref Full Text | Google Scholar

Damore, D. F. (2005). Issue convergence in presidential campaigns. Polit. Behav. 27, 71–97. doi: 10.1007/s11109-005-3077-6

Crossref Full Text | Google Scholar

De la Torre, C., and Conaghan, C. (2009). The hybrid campaign: tradition and modernity in Ecuador's 2006 presidential election. Int. J. Press/Politics 14, 335–352. doi: 10.1177/1940161209334523

Crossref Full Text | Google Scholar

Downs, A. (1957). An economic theory of democracy. New York: Harper & Row.

Google Scholar

Espinoza Bianchini, G., Navia, P., and Ulriksen Lira, C. (2021). Using online social networks to acquire political information: the politically engaged non-ideological youth in Chile, 2017–2019. Int. J. Politics Cult. Soc. 35, 497–515. doi: 10.1007/s10767-021-09407-6

Crossref Full Text | Google Scholar

Fagan, E. J. (2021). Issue ownership and the priorities of party elites in the United States, 2004–2016. Party Polit. 27, 149–160. doi: 10.1177/1354068819839212

Crossref Full Text | Google Scholar

Feierherd, G., and Lucardi, A. (2022). When the partisan becomes personal: mayoral incumbency effects in Buenos Aires, 1983–2019. J. Elections Public Opin. Parties. 33, 684–704. doi: 10.1080/17457289.2022.2081696

Crossref Full Text | Google Scholar

Fowler, E. F., Franz, M. M., Martin, G. J., Peskowitz, Z., and Ridout, T. N. (2021). Political advertising online and offline. Am. Polit. Sci. Rev. 115, 130–149. doi: 10.1017/S0003055420000696

Crossref Full Text | Google Scholar

Fowler, E. F., Ridout, T. N., and Franz, M. M. (2016). “Political advertising in 2016: the presidential election as outlier?” The Forum

Google Scholar

Franz, M. M., Fowler, E. F., Ridout, T., and Wang, M. Y. (2020). The issue focus of online and television advertising in the 2016 presidential campaign. Am. Politics Res. 48, 175–196. doi: 10.1177/1532673X19875722

PubMed Abstract | Crossref Full Text | Google Scholar

García Rodríguez, J. I. (2013). “Las franjas electorales en la experiencia chilena.” (16): 279–293.

Google Scholar

Gelman, A., and Imbens, G. (2019). Why high-order polynomials should not be used in regression discontinuity designs. J. Bus. Econ. Stat. 37, 447–456. doi: 10.1080/07350015.2017.1366909

Crossref Full Text | Google Scholar

Goggin, S. N., Henderson, J. A., and Theodoridis, A. G. (2020). What goes with red and blue? Mapping partisan and ideological associations in the minds of voters. Polit. Behav. 42, 985–1013. doi: 10.1007/s11109-018-09525-6

PubMed Abstract | Crossref Full Text | Google Scholar

Graber, D. (2001). Mass media and American politics. Washington: CQ Press.

Google Scholar

Greene, K. F., and Baker, A. (2011). The Latin American Left's mandate: free-market policies and issue voting in new democracies. World Polit. 63, 43–77. doi: 10.1017/S0043887110000286

PubMed Abstract | Crossref Full Text | Google Scholar

Grofman, B. (2004). Downs and two-party convergence. Annu. Rev. Polit. Sci. 7, 25–46. doi: 10.1146/annurev.polisci.7.012003.104711

Crossref Full Text | Google Scholar

Imbens, G., and Kalyanaraman, K. (2012). Optimal bandwidth choice for the regression discontinuity estimator. Rev. Econ. Stud. 79, 933–959. doi: 10.1093/restud/rdr043

PubMed Abstract | Crossref Full Text | Google Scholar

Imbens, G. W., and Lemieux, T. (2008). Regression discontinuity designs: a guide to practice. J. Econ. 142, 615–635. doi: 10.1016/j.jeconom.2007.05.001

Crossref Full Text | Google Scholar

Kaid, L. L., and Johnston, A. (1991). Negative versus positive television advertising in US presidential campaigns, 1960-1988. J. Commun. 41, 53–64. doi: 10.1111/j.1460-2466.1991.tb02323.x

Crossref Full Text | Google Scholar

Karlsen, R., and Aardal, B. (2016). Political values count but issue ownership decides? How stable and dynamic factors influence party set and vote choice in multiparty systems. Int. Polit. Sci. Rev. 37, 261–276. doi: 10.1177/0192512114558456

Crossref Full Text | Google Scholar

Lee, D. S. (2008). Randomized experiments from non-random selection in US house elections. J. Econ. 142, 675–697. doi: 10.1016/j.jeconom.2007.05.004

Crossref Full Text | Google Scholar

Lee, D. S., and Lemieux, T. (2010). Regression discontinuity designs in economics. J. Econ. Lit. 48, 281–355. doi: 10.1257/jel.48.2.281

Crossref Full Text | Google Scholar

Lisbona, F., and Navia, P. (2018). Los mensajes de la propaganda televisiva en las campañas presidenciales de Michelle Bachelet y Sebastián Piñera en Chile, 2005-2013. Rev. Mex. Cienc. Polit. Soc. 63, 333–363. doi: 10.22201/fcpys.2448492xe.2018.233.63010

Crossref Full Text | Google Scholar

López-Hermida, A. (2009). La imagen de la mujer en política: La campaña electoral televisiva de Michelle Bachelet. Cuadernos. Info 24, 9–16. doi: 10.7764/cdi.24.30

Crossref Full Text | Google Scholar

López-Hermida Russo, A. P. (2017). “Political advertising in Chile” in Routledge handbook of political advertising. eds. C. Holtz-Bacha and M. R. Just (New York: Routledge), 101–112.

Google Scholar

Morales, M. (2008). La primera mujer presidenta de Chile. ¿Que explicó el triunfo de Michelle Bachelet en las elecciones de 2005-2006? Lat. Am. Res. Rev. 43, 7–32. doi: 10.1353/lar.2008.0013

PubMed Abstract | Crossref Full Text | Google Scholar

Nickelsburg, M., and Nortpoth, H. (2000). Commander-in-chief or chief economist? The president in the eye of the public. Elect. Stud. 19, 313–332. doi: 10.1016/S0261-3794(99)00054-2

Crossref Full Text | Google Scholar

Petrocik, J. R. (1996). Issue ownership in presidential elections, with a 1980 case study. Am. J. Polit. Sci. 40, 825–850. doi: 10.2307/2111797

PubMed Abstract | Crossref Full Text | Google Scholar

Petrocik, J. R., Benoit, W. L., and Hansen, G. J. (2003). Issue ownership and presidential campaigning, 1952-2000. Polit. Sci. Q. 118, 599–626. doi: 10.1002/j.1538-165X.2003.tb00407.x

PubMed Abstract | Crossref Full Text | Google Scholar

Pons, V., and Tricaud, C. (2018). Expressive voting and its cost: evidence from runoffs with two or three candidates. Econometrica 86, 1621–1649. doi: 10.3982/ECTA15373

Crossref Full Text | Google Scholar

Porath, W., Suzuki, J. J., Ramdoor, T. M., and Portales, J. C. (2015). Newspaper coverage of three presidential campaigns in Chile: personalisation and political strategies. Bull. Lat. Am. Res. 34, 451–466. doi: 10.1111/blar.12366

Crossref Full Text | Google Scholar

Portales, J. C. (2009). A test of personalization of politics in Chile’s 2009 presidential election. Cuadernos. Info 25, 69–82. doi: 10.7764/cdi.25.48

Crossref Full Text | Google Scholar

Posner, P. W. (2003). Local democracy and popular participation: Chile and Brazil in comparative perspective. Democratization 10, 39–67. doi: 10.1080/13510340312331293927

Crossref Full Text | Google Scholar

Romer, T., and Rosenthal, H. (1979). The elusive median voter. J. Public Econ. 12, 143–170. doi: 10.1016/0047-2727(79)90010-0

Crossref Full Text | Google Scholar

Scherman, A., and Arriagada, A. (2012). Disposition to vote and media consumption patterns among Chilean youth. Comun. Soc. 25, 85–112. doi: 10.15581/003.25.36175

Crossref Full Text | Google Scholar

Shadish, W. R., Cook, T. D., and Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Boston, MA: Houghton, Mifflin and Company.

Google Scholar

Sides, J. (2006). The origins of campaign agendas. Br. J. Polit. Sci. 36, 407–436. doi: 10.1017/S000305542100112X

Crossref Full Text | Google Scholar

Sides, J., Vavreck, L., and Warshaw, C. (2022). The effect of television advertising in United States elections. Am. Polit. Sci. Rev. 116, 702–718.

Google Scholar

Simon, A. F. (2002). The winning message: Candidate behavior, campaign discourse, and democracy. Cambridge: Cambridge University Press.

Google Scholar

Stubager, R., and Slothuus, R. (2013). What are the sources of political parties’ issue ownership? Testing four explanations at the individual level. Polit. Behav. 35, 567–588. doi: 10.1007/s11109-012-9204-2

Crossref Full Text | Google Scholar

Therriault, A. (2015). Whose issue is it anyway? A new look at the meaning and measurement of issue ownership. Br. J. Polit. Sci. 45, 929–938. doi: 10.1017/S000712341400057X

Crossref Full Text | Google Scholar

Uribe, R., Buzeta, C., Manzur, E., and Pefaur, N. (2018). Desgastada pero aún efectiva: examinando los datos de audiencia de la franja electoral presidencial chilena (1999-2017). Cuadernos. Info 43, 181–199. doi: 10.7764/cdi.43.1223

Crossref Full Text | Google Scholar

Valenzuela, S., and Arriagada, A. (2011). Politics without citizens? Public opinion, television news, the president, and real-world factors in Chile, 2000-2005. Int. J. Press/Polit. 16, 357–381. doi: 10.1177/1940161210379636

Crossref Full Text | Google Scholar

Walgrave, S., and Lefevere, J. (2017). Long-term associative issue ownership change: a panel study in Belgium. J. Elections Public Opin. Parties 27, 484–502. doi: 10.1080/17457289.2017.1285305

Crossref Full Text | Google Scholar

Walgrave, S., Lefevere, J., and Tresch, A. (2012). The associative dimension of issue ownership. Public Opin. Q. 76, 771–782.

Google Scholar

Walgrave, S., Tresch, A., and Lefevere, J. (2015). The conceptualisation and measurement of issue ownership. West Eur. Polit. 38, 778–796. doi: 10.1080/01402382.2015.1039381

Crossref Full Text | Google Scholar