Do Candidate Debates Matter?  An International Perspective

Within the US election cycle, there are some pundits who believe that “debates don’t matter;” that they don’t have an impact on an electorate that has already made up their minds. However, NDI’s overseas work in establishing and institutionalizing Candidate Debates suggests otherwise. NDI’s programming not only supports candidate debates, but seeks to identify and measure the outcomes and impacts of debates on emergent democracies. To that end, and with the support of the National Endowment for Democracy (NED), NDI’s Monitoring, Evaluation & Learning team commissioned an Evidence Scoping Paper on Candidate Debates as a way to more systematically engage with academic researchers, outline a Debates Outcome Model, and develop measurement strategies to better understand the dynamics of candidate debates across different contexts and political conditions.

NDI's Director of Monitoring and Evaluation Linda Stern spoke with with Eric Kramon, Associate Professor of Political Science at George Washington University and NDI's Global Advisor for Candidate Debate Programs Matt Dippell to discuss the role of debates in the democratic development process overseas. 

Linda: Let me start with you Matt. Can you talk about NDI’s support for Candidate Debates and how they fit into NDI’s international democracy assistance portfolio? 

Matt: NDI is increasingly seeing countries around the globe make candidate debates a cornerstone of their elections. By our count, more than 90 countries to date have held debates. Driving this international trend are the many benefits that debates bring to established and emerging democracies, including helping voters make informed choices at the ballot box; focusing candidates on policy issues rather than personality, religion or ethnic loyalties; reducing election tensions or the potential for violence in countries emerging from conflict, and holding elected officials to their campaign promises. NDI has teamed with the Commission on Presidential Debates (CPD), the non-profit, non- partisan organization that has produced and sponsored all 33 U.S. general election presidential and vice presidential debates since 1987, to help sponsors stage debates for the first time or improve on past forums. NDI and the CPD have collectively helped partner groups organize more than 400 debates in some 45 countries for candidates ranging from mayors and legislators to prime ministers and presidents. Together with other debate groups, NDI and the CPD co-founded a 38-country international network of debate organizations and a complementary resource website, both known as Debates International, to foster the exchange of experiences to improve debates.

Linda: And how do you see debates as a complement to NDI’s other programmatic areas?

Matt:  In practice,  debates are much more than a 2-hour event. For some of the reasons just mentioned, by engaging often millions of viewers and listeners, debates can serve as a catalyst for a range of democracy strengthening activities that multiply their impact. Complementary activities can include debate watch parties that bring citizens of different political perspectives together to discuss issues, as well as civic education campaigns that inform voters on the positions of candidates and encourage them to go out and vote. Preparing for debates can also help motivate candidates and political parties to develop responsive policy platforms and distill their key messages to better connect with voters. Debates also put candidates on the public record so that the media and civic advocacy groups can hold them accountable to campaign commitments once in office. 

Linda: Eric, let’s start with this idea that “debates don’t matter;” the idea that by the time voters listen to debates, their minds are already made up. Can you talk about what you found in Ghana and from the evaluation of NDI’s debates program in Malawi that suggest otherwise?

Eric:   In my collaborative work in Ghana with Sarah Brierley and George Ofosu and also in my work with NDI in Malawi, we’ve looked at this exact question of whether debates matter. It’s helpful to start by asking, what is it about debates that we think is going to have an effect on people? Typically, those of us who study this think along two dimensions. One is that people are going to learn about the policy positions of candidates. A second is that voters learn other things about candidates by seeing them on the debate stage, engaging with other candidates. So they might learn about their qualifications, about  their character, or their trustworthiness, just by having the opportunity to see or hear them in the debate format, which is different from the usual ways people often encounter candidates.

In the research, we find that people do learn a lot about these dimensions. In Malawi, we asked people afterwards, “what was the most important thing they took away from the debate?” Many people said,  “this is the first time I’ve heard the policy platforms of the parties” or “Now I know more about the policy platform of such and such party.” This also shows up in survey data when we compare the people who watched the debates with people who don’t, and we find that debate watchers can report more accurately the party platforms on issues discussed in the debates. 

We also ask, how does this learning translate into voter preferences or voting choices? There are a lot of common trends that hold in Ghana and Malawi, but also in some of the other places where debates have been studied. One is that debates tend to improve people’s views of challengers or less well-known candidates, because these are people they know less about and learn more about through the debates.  There is also some evidence on policy based voting: In Malawi we found that those who watched the debates were more likely to vote for candidates who shared their policy preferences; so once they learned what the policy platforms were, people moved toward candidates that shared their views. And then one of the most interesting results that has been found in almost all studies on debates--outside of the US context--is that they tend to make people more open to candidates across partisan lines. In the Ghana and Malawi case, we found evidence that people were more likely to vote across party lines after watching the debate, so if you’re a strong supporter of one party, you watch the debate, there’s a greater likelihood that you will vote across party lines. But beyond vote choice, we’ve found that the debates  reduce polarization. So even if it doesn’t change how people vote, the difference in how they evaluate the candidates is diminished. And we know that polarization poses all sorts of challenges for democracy.So to the extent that debates can help people understand opposing party policy positions a bit better and introduce people to opposing candidates--even if their votes ultimately do not change-- they can still play a helpful role in reducing polarization.  

Finally, it is important to note that what I am describing here are average impacts.  A lot of the impact of debates of course depends on the candidates themselves and how they conduct themselves.

Linda: And what about the staying power of those changes in preferences across partisan lines? How long do those shifts in preference last?

Eric: Yes, here we have some mixed findings. In Ghana,some of the immediate effects of the debates dissipated fairly quickly, mainly in places where people were living in dominant strongholds of one of the parties. So they watched the debate, they kind of become more open to the opposing party, but then they go straight back into a social environment that isn’t very friendly to that party, so the effects go away. But people who are living in more competitive places, where people are exposed to more diverse political views, there’s more staying power. In the Malawi case, the immediate effects on polarization do dissipate over time--other aspects of the campaign also have an impact--but they don’t disappear altogether.

Linda: Eric, from scoping the evidence, what are some other potential outcomes of candidate debates for emerging democracies? 

Eric: There are a number of things that we’ve looked at and should be looking at going forward. We can look at other forms of engagement, political participation--so do debates motivate people to get more active in the campaigns? A very interesting potential thing that has been looked at in Sierra Leone is how candidates campaign after they participate in debates. So do they change how they talk about policy? Does it make them more likely to engage with voters on policy, now that they’ve had the chance to hone their policy arguments? Another dimension would be politician behavior after they’re in office. So they’ve just made these promises in the debate, there’s more common knowledge around what they said they’re going to do--does this make it more likely that they’re going to follow through? Or if they don’t follow through, does it make it more likely that they’re going to face some form of accountability for that once they’re in office? And then more generally,  there are potential outcomes related to democratic norms, general political tolerance, effects on tensions between different partisan groups. Even if debates don’t affect how particular people vote, they could have consequences for these broader aspects of the political system.

Linda: So Matt, the emerging evidence suggests that the strongest effects of debates are on voter knowledge.  From a practitioner perspective, what are the elements within Candidate Debates that are different from other forms of Civic or Voter Education that also seek to impact knowledge gain? 

Matt: One of the reasons that debates are emerging as an expected and indispensable part of elections is because they are in many ways unique events. Debates are generally the only time in the course of a political campaign where the candidates come together in a neutral setting that  allows citizens to make a side by side comparison of their positions and leadership styles. Debates are really  the ultimate job interview for elected officials. Absent debates, in some countries voters are left to assess candidates and their policies based on  political advertisements or party rallies, which do not provide the same comparative information or insights into candidates. At a debate, the moderator or fellow debater poses questions and follow-ups that require candidates to provide in-depth responses that go  beyond slogans. Candidates also have the invaluable opportunity to speak directly and unfiltered to the voters.   

Linda: Matt, Eric’s research confirms  that debates do matter; they can change voter preferences and behaviors, and potentially strengthen electoral accountability, even after elections. However, what have you seen as some of the challenges in supporting partners organizing candidate debate programs?

Matt:. Despite differences in political systems and culture, debates in some countries have failed when sponsors could not surmount common challenges, including allaying fears that organizers may have a political bias, convincing reluctant candidates to participate, negotiating with rival media outlets to show a common debate broadcast and successfully producing live national television and radio programs, among other issues. Of these, securing the participation of candidates is the most universal hurdle. In this regard, the main means of convincing candidates to take part is to help give voice to public expectations and demand for debates with the help of the media and civic advocacy groups. NDI has found that polls in countries ranging from Kenya to Peru show that an average of 75% of citizens want to see their candidates debate or risk paying a cost at the ballot box for disappointing voters by not taking part. It is notable that in a handful of countries, debates are seen as so important to elections that laws have been passed to require candidates to debate -- although public expectations continue to be a bigger motivator to participate than the threat of legal sanctions such as the loss of public airtime.

Linda: Eric, what does the research say in terms of overcoming some of these challenges and incumbent disincentives? 

Eric: I think the research speaks to this in a few ways. First,it shows pretty clearly that voters don’t like it when candidates, especially the incumbent, don’t show up. So in the Malawi study, the incumbent did not show up, and we can document really large drops in favorability toward the incumbent, toward intended vote choice for the incumbent, immediately after that. And a lot of that does persist through election day, although it dissipates, in part because of how much incumbency advantage there is in Malawi. One of the interesting things was that the main people who were not happy with the president’s decision not to debate in Malawi were actually his own party supporters. So it had the effect of making part of his core vote bloc more interested or open toward the challenger candidates, which was really interesting. So one incentive is to show up is that the voters don’t like it if you do not..One study in Liberia, by Horacio Larreguy and Jeremy Bowles, looked at different ways of contacting,reminding, and encouraging candidates to show up (this is in the context of legislative elections) in Liberia. And they did find that sending emails and text messages from respected journalists did increase the likelihood that candidates--especially incumbents--would come to the debates. 

Linda: So Matt and Eric, the emerging evidence suggests that the policy component of debates might be the driver of impact on voters, at least in terms of knowledge gain. This is happy news for practitioners. From a research and practitioner perspective, what are some additional areas of research and evaluation that would be important to pursue?

Matt:  Additional research would help debate organizers quantify and improve their efforts in a range of areas. Pending topics could include: What formats are most effective in aiding voters to learn about candidates? What is the cost-effectiveness of debates as a program intervention? Given that debates reach millions of citizens, it would seem hard to find other voter education activities that would match the impact. How large a role do debates play in creating political space in closing societies? In many countries, debates are viewed as an essential democratic practice in their own right. What are the cumulative effects of debates overtime?

Eric: I think there are two broad areas: one would be moving beyond focusing on preferences and vote choices and into broader outcomes related to democratic norms and accountability. And generally broader effects on the democratic system and process. And a second would be looking more systematically at what happens as debates become less new. A lot of research studies have been conducted- in places where debates are relatively new, and so we would want to see what happens over time as they become more routine aspects of the electoral process.

Linda: Thank you both for your insights and expertise. We forward to working with both of you to pilot measurement strategies for this important facet of democratic development.