Inter-temporal Evidence on the Strategy of Populism

• In Event: The Political Economy of Radical and Populist Parties

Abstract

Campaigns are essential components of party competition and crucial to the functioning of democratic politics. For this reason, many scholars across the social sciences devote a large attention to political campaigns. One of the main goal in an electoral campaign is to influence the behavior of prospective voters, persuading them to show up at the polls and to make the right choice. Rhetoric can be seen as a tool that candidates use to persuade or mobilize voters to their advantage, by strategically emphasizing issues in voters’ minds (Sides 2006).

In this paper, we investigate whether political candidates use populist rhetoric strategically, i.e. if they try to match the electorate’s demand for populism by increasing or decreasing the supply during their electoral campaign. We do so by looking, first, at the 2016 presidential campaign and, second, at the 2018 midterm electoral campaign in the United States. In the first exercise,
we observe how the each of the two presidential candidates adapts the level of populism according to the characteristics of the audience. In this respect, we construct a measure of populism that varies within candidates and across speeches. The second exercise allows us to generalize the suggestive evidence from the presidential campaign to a larger sample of candidates; here we explore how the variation of the level of populism across candidates can be (partly) explained by features that we detected in the Presidential campaign.

We use automated text analysis to create a continuous index of populist rhetoric at the speech or candidate level. In particular, we exploit a dictionary of populist words developed in political science (Pauwels 2011, Rooduijn & Pauwels 2011) to create an intensity measure based on word frequency. The use of this dictionary has two main advantages: it captures the core of different definitions of populism, i.e. the juxtaposition of a corrupt elite with a morally virtuous people (Mudde, 2017), and it allows the creation of a speech-level intensity measure of populism. This operationalization treats populism first and foremost as a rhetoric tool, rather than as an ideology.

The literature on populism points to the role played by economic insecurity in generating demand for populism. Hence, we augment our dataset on candidates with a set of variables that capture different aspects of economic insecurity. Inspired by the literature, we adopt three main proxies: (i) the increase in inequality, (ii) the drop in manufacturing employment and (iii) perceptions of economic insecurity. Further, if the supply of populism is strategic, we expect this
to vary according to local political conditions. We explore the role of
(i) the competitiveness of the race and (ii) the degree of polarization of electorate.

We have 3 distinct sets of results. First, we detect marked differences between presidential candidates in the last and previous elections back to 2004. Second, we identify some characteristics that can be credibly associated to
populist candidates, and verify that these features can be detected also among the mid-term election candidates. Third, we demonstrate that the supply of populist rhetoric is tailored on socioeconomic characteristics of the rally audience,
suggesting strategic behavior on the supply of populism.

More specifically, we find that Donald Trump is significantly more populist than his competitor and his predecessors. Moreover, Trump seems to be more strategic in the use of populist rhetoric; he is more ”populist” in swing states, in less polarized districts, and in those MSAs characterized by more economic uncertainty. We then analyze the behavior of all the congressional candidates of the traditional parties in the mid-term congressional elections of 2018 to detect analogies with the 2016 presidential campaign. We provide evidence that individual characteristics of the congressional candidates are relevant drivers for the use of populist rhetoric. Specifically we find that incumbent and old candidates are significantly associated with lower level of populism, whilst outsider candidates are associated with higher level of populism. Besides, our results strongly suggest that outsiders with higher educational attainment use populist rhetoric strategically, responding to the same drivers identified in the Presidential campaign.

Our results suggest that, more than an intrinsic individual or party characteristic, populism is a rhetoric tool to which some candidates can strategically recur to mobilize or convince voters in their favor.

Authors

• Giampaolo Lecce

• Gloria Gennaro, Bocconi

• Massimo Morelli, Bocconi University