Making rallies great again
The rise of populism in recent decades has raised concerns about the future of liberal democracy, especially in the US and Western Europe (Hawkins et al. 2019). The vote share of populist parties has tripled in Europe from 1998 to 2018 (Lewis et al. 2018). The success of these parties is often attributed to individual leaders, figures such as Jörg Haider in Austria, Matteo Salvini in Italy, Recep Tayyip Erdogan in Turkey, and Narendra Modi in India. Commentators have even coined terms such as the ‘Trump effect’, the ‘Le Pen effect’, and the ‘Haider phenomenon’ to highlight their importance.
One campaign activity that is especially important for many populist leaders is holding large rallies (Jansen 2011). These rallies help populists appeal directly to ‘the people’ and to promote the idea that they are particularly close the people (Albertazzi and McDonnell 2007, Mudde and Kaltwasser 2017). They also help populist parties and movements attract media attention and spread their message more broadly across the population. Rallies may also “offer an emotive substitute for substantial political organization and engagement” (Lichtenstein 2019).
Rallies played a prominent role in the 2006 campaign of Rafael Correa (populist president of Ecuador from 2007 to 2017); Erdogan held more than 100 rallies during his month-long campaign leading up to the 2019 elections in Turkey; and Rodrigo Duterte’s rallies contribute to his large grassroots support in the Philippines, just to give a few examples (de la Torre 2019, Curato 2016).
In recent research (Snyder and Yousaf 2020), we provide evidence that populist leaders may be particularly effective in gaining support via their campaign rallies, at least temporarily. We do this by studying the effect of campaign rallies held by each of the Democratic and Republican US Presidential candidates since 2008, including Donald Trump. Specifically, we study the impact of rallies on citizens’ preferences over candidates, policy issues, and their intention to vote.
Donald Trump and populism
While the Republican Party is not easily characterised as populist, the current Republican US President, Donald Trump, is. Hawkins et al. (2019) analyse the speeches of 215 world leaders around the world and find that Donald Trump is a populist leader with a score of 0.78 (87th percentile). The previous two presidents, Barack Obama and George W. Bush, have scores of 0.15 and 0.21, respectively.
Further evidence of Trump’s populistic style and appeal can be seen by examining the many ways his 2016 presidential campaign and unlikely victory represent a break from the past. Trump won the Republican nomination even though he was opposed by virtually all Republican elites (MacWilliams 2016), and spent much less on his campaign than the other leading contenders (National Public Radio 2016). In popular media, Trump was widely referred to as a populist. Observers often draw comparisons between Trump and other nationalistic and anti-immigrant populist leaders, such as Matteo Salvini in Italy, Marine Le Pen in France, or Hungarian Prime Minister Viktor Orbán.
Measuring the impact of political rallies
Assessing the impact of rallies by presidential candidates is difficult for two main reasons. First, presidential candidates tend to target areas located in swing states. For example, in 2016, between 22 September and election day (8 November) in 2016, only two out of the 76 rallies by Donald Trump, and two out of the 58 rallies by Hillary Clinton, were in non-battleground states. Second, to uncover the impact of rallies on political decisions, we need a measure of political preferences for different areas, both before and after a political rally.
To overcome these potential concerns, we use data from a large survey – the Cooperative Congressional Elections Study – to exploit both time and geography. Some respondents in the Study were surveyed a few days before a rally, while others were surveyed a few days afterwards. For many rallies, some of the respondents surveyed around the time of the rally lived near the rally site.
Thus, we compare the average change in political preferences and voting intentions around the time of the rally among respondents who live near a rally site, with the changes among those who live far away from the site.
Since the 2016 Congressional Elections Study has a total sample size of almost 65,000, and since Trump held 71 rallies in 15 states during the survey period, mostly at sites in large media markets, even after restricting the sample we have many thousand respondents.
Our findings are straightforward. First, we find that Trump rallies appear to have increased support for him over Clinton by about 4.5 percentage points on average (Figure 1, panel a). The effects are short-lived, lasting about one week. Moreover, the effects are especially large for respondents who identify as ‘weak’ Republicans and Republican-leaning independents. Second, Trump rallies also appear to have increased the intention to turn out for ‘strong’ Republicans and Democrats, but not for other respondents (Figure 1, panel b). Third, rallies by Trump increased the individual campaign contributions for him (Figure 1, panel c). The effect is again short-lived, dying out in less than six days.
Figure 1 Effect of rallies on support for Trump
(a) Intention to vote for Trump
(b) Intention to go vote (turn out)
(c) Individual campaign contributions to Trump
To compare how the effect of rallies by Trump differs from rallies by other presidential candidates, we use the same approach to study the rallies held by Hillary Clinton in 2016, as well as those held by Barack Obama and Mitt Romney in 2012, and by Obama and John McCain in 2008. For these candidates – none of whom could be called a populist – we find no consistent, robust effects.
It is intriguing that rallies by Trump but no other past candidate led to a change in political preferences among the electorate. To explore the channels that may explain these findings, we first analyse the local media coverage of rallies around the time these rallies were held.
We find evidence that Trump’s local media coverage increased around the time of his rallies, both on television news and in newspapers. We do not find similar effects on the amount of paid television advertising by either Trump or Clinton. So, it is possible that respondents changed their vote intentions due to the additional coverage of the candidate on media.
However, we find increases of similar magnitude in the coverage of almost all other presidential candidates around their rallies as well. This suggests that either the quality of Trump’s coverage was different from that of other candidates, or that the effect of the additional coverage on respondents was different.
Another channel through which presidential rallies may impact voters is by changing voters’ preferences on certain policies. A candidate’s rallies might convince some voters that the policy platform proposed by the candidate is better than that of their opponent, which may shift voters’ preferences closer to those of the candidate. We find that rallies by Trump did not lead to a change in the voters’ preferences over issues and voters’ beliefs about the relative importance of issues.
Although we lack measures to investigate this, it is possible that Trump’s rallies were more entertaining than typical political rallies and grabbed people’s attention in a way that other candidates’ rallies did not. De la Torre (2019) argues that “populism blurs the line between politics and entertainment”. Lemieux (2020) recently wrote, “What is said and done in President Trump’s rallies is more entertaining than information on a political programme. These rallies are liturgical spectacles of fusion between the great leader and ‘the people’.”
We find evidence that populist leaders may be particularly effective in gaining support via their campaign rallies, at least temporarily. Trump’s 2016 rallies – unlike those of other recent US presidential candidates – led to an increase in the number of people intending to vote for him, the number intending to turn out, and the amount of individual contributions to his campaign.
These results provide suggestive evidence that rallies are a more important tool for more populist candidates. Populist leaders’ success may depend on connecting with voters via rallies. Since the effects seem short-lived, these findings might also help account for the fact that many populist leaders hold rallies frequently and not only during the months leading up to elections.
Albertazzi, D, and D McDonnell (2007), Twenty-ﬁrst century populism: The spectre of Western European democracy, Springer.
Curato, N (2016), “Flirting with authoritarian fantasies? Rodrigo Duterte and the new terms of Philippine populism”, Journal of Contemporary Asia 47(1): 142–53.
De la Torre, C (2019), Populisms: A quick immersion, Tibidabo Publishing.
Hawkins, K A, R Aguilar, B Castanho Silva, E K Jenne, B Kocijan and C Rovira Kaltwasser (2019), “Global populism database, v1”, Harvard Dataverse, V1.
Jansen, R S (2011), “Populist mobilization: A new theoretical approach to populism”, Sociological Theory 29(2): 75–96.
Lemeiux, P (2020), “From democracy to populist rallies”, The Library of Economics and Liberty.
Lewis, P, S Clarke, C Barr, J Holder and N Kommenda (2018), “The new populism”, The Guardian, 20 November.
Lichtenstein, N (2019), “Are trade unions and their members ‘populist’?”, IdeAs: Idées d’Amériques 14.
MacWilliams, Matthew C (2016), “Who decides when the party doesn’t? Authoritarian voters and the rise of Donald Trump”, Political Science and Politics 49(4): 716–21.
Mudde, C, and C Rovira Kaltwasser (2017), Populism: A very short introduction, Oxford: Oxford University Press.
Snyder, Jr, J and H Yousaf (2020), “Making Rallies Great Again: The Effects of Presidential Campaign Rallies on Voter Behavior, 2008-2016”, NBER Working Paper 28043.