EXISTING APPROACHES

Although this special issue focuses on the ideational approach, the literature on populism and liberal democracy draws heavily from a second, competing approach: the political–strategic approach. Without considering both approaches, it can be difficult to understand the design and the implications of current studies.Footnote ²

The ideational approach defines populism as a thin-centered ideology or a discourse that frames politics as a struggle between the will of the common people and an evil, conspiring elite (Hawkins et al. Reference Hawkins, Carlin, Littvay and Kaltwasser2019). This approach tends to see populist incumbents having two types of impact. On the one hand, leaders who use this rhetoric may have a positive impact on democratic participation and representation because of their efforts to reshape the political agenda and to build new institutions of democratic inclusion (Canovan Reference Canovan1999; Mény and Surel Reference Mény and Surel2002). Populist incumbents make these efforts because of their belief in popular sovereignty and because populist forces are arising in response to actual failures of liberal democracy to fully represent the electorate. In this way, the emergence of populism can be a corrective for democracy (Mudde and Rovira Kaltwasser Reference Mudde and Kaltwasser2012).

On the other hand, populist incumbents consistently harm democratic contestation and other core institutions of liberal democracy (Abts and Rummens Reference Abts and Rummens2007; Canovan Reference Canovan1999; Mény and Surel Reference Mény and Surel2002; Mudde and Rovira Kaltwasser Reference Mudde and Kaltwasser2012; Urbinati Reference Urbinati2019). Generally, the literature highlights three mechanisms. First, because populist leaders and their followers perceive democracy as the expression of what ordinary people would want if they had unbiased information and could think freely of elite influence—a popular will that populists believe they know and alone embody—populist leaders are more willing than non-populists to violate norms and rules of electoral competition once they are in office (Abts and Rummens Reference Abts and Rummens2007, Reference Abts and Rummens417; Müller Reference Müller2016; Ochoa Espejo Reference Ochoa Espejo and de la Torre2015, Reference Ochoa Espejo and de la Torre61). Second, because populists see their opponents as powerful agents who knowingly conspire against the common good, they believe they must curtail their civil liberties and make them pay for their crimes, even at the expense of due process. Third, because electorally successful populist forces are often personalistic movements (Pappas Reference Pappas2016), followers are frequently motivated to concentrate government powers in the hands of their chief executive, whom they see as the embodiment of their will. Thus, populist incumbents undermine horizontal accountability.

In contrast to the ideational approach, the political–strategic approach views populist incumbents as having a more consistently negative impact on liberal democracy—indeed, on the same institutions mentioned by ideational theorists. This approach explains this negative impact not as a function of populist ideas but instead by what it sees as the defining features of populist forces: a personalistic organization led by charismatic political outsiders (Barr Reference Barr2017; Levitsky and Loxton Reference Levitsky and Loxton2013). Again, three mechanisms are posited. First, because populists are often political outsiders, they experience high costs in building relationships with traditional political elites. Hence, once in power, they prefer to undermine the party system and rebuild political institutions from the constitution up (Levitsky and Loxton Reference Levitsky and Loxton2013). Second, because populist forces are personalistic movements, once in power, their leaders find it easier and simpler to dismantle checks and balances in favor of the executive branch, and they need control of the media to communicate with their followers (Kenny Reference Kenny2020; Weyland Reference Weyland2024). Third, as vote-maximizing politicians, populist incumbents are keen on satisfying constituents who voted them into office on a platform of radical institutional change and “cleaning house.” Indeed, the support of large numbers of voters gives them the power to carry out this mandate (Levitsky and Loxton Reference Levitsky and Loxton2013).

RESULTS: POPULISM AS THREAT

In recent years, empirical studies from both approaches have consistently confirmed the broad claims of a negative association between populism and democratic contestation/liberal democracy. Early research on the impact of populism on democracy often featured individual country or small-N studies (e.g., Betz Reference Betz2001; de la Torre Reference de la Torre1997; Hawkins Reference Hawkins2003). However, the literature made an important step forward with the publication of Mudde and Rovira Kaltwasser’s (Reference Mudde and Kaltwasser2012) book, which not only included theoretical contributions from the ideational side but also was one of the first to compare populism across regions. Contributors found that contemporary populists in Europe and Latin America were associated with declines in democratic contestation, although these effects were moderated by the prior strength of democratic institutions. Another early comparative study from the political–strategic approach was Levitsky and Loxton’s (Reference Levitsky and Loxton2013) article on populism and democracy in Latin America. In an analysis of several contemporary governments in the region, they argued that populist outsiders were associated with significant declines in electoral quality, civil liberties of opponents, and horizontal accountability. It is important that they noted the primacy of elections and plebiscites, which grant populists the legitimacy to weaken democratic institutions, ultimately leading to comparative authoritarian regimes. Similar findings emerged in a study of Latin America by de la Torre and Arnson (Reference de la Torre and Arnson2013).

Since 2016, researchers have confirmed these findings with larger datasets using broad indicators of democratic contestation. In two seminal large-N studies, Huber and Schimpf (Reference Huber and Schimpf2016a, Reference Huber and Schimpf2016b) found that populist incumbents were associated with overall declines in democracy (measured in terms of the Polity IV polity2 measure) in both Latin America and Europe. Three cross-regional studies—Kyle and Mounk (Polity IV, 2018), Juon and Bochsler (Democracy Barometer, 2020), and Ruth-Lovell and Grahn (Varieties of Democracy, 2023)—also found that populism was associated with declines in liberal democracy. Most of these studies based their hypotheses on the ideational approach.

Other research focused on specific institutions of liberal democracy, including media freedom, the rule of law, and the strength of the constraints placed on the executive’s powers (i.e., horizontal accountability). In this instance, a few studies (Kenny Reference Kenny2017, Reference Kenny2020) drew more clearly from the political–strategic approach. Huber and Schimpf (Reference Huber and Schimpf2017) found that populism in Europe was particularly harmful to horizontal accountability. Houle and Kenny (Reference Houle and Kenny2018) followed this finding with a study showing that populism was associated with declines in horizontal accountability, judicial freedom, and the rule of law in Latin America, which Kenny (Reference Kenny2017) confirmed with a global dataset. In a Latin American study, Ruth (Reference Ruth2018) showed the conditions under which populism was more likely to lead to declines in horizontal accountability. Finally, Kenny (Reference Kenny2020) extended his 2017 analysis with a cross-regional study on the negative impact on media freedom. It is important to note that many of these studies demonstrate that these negative effects are general to populism of the left and right (e.g., Huber and Schimpf Reference Huber and Schimpf2017; Juon and Bochsler Reference Juon and Bochsler2020; Kyle and Mounk Reference Kyle and Mounk2018; Vittori Reference Vittori2022).

Some of the most recent literature focuses on the varying effects of populist incumbents on contestation. Scholars recognize that our tendency to select on the dependent variable has given the illusion that all populist governments lead to backsliding. However, this is not quite true. Weyland (Reference Weyland2024) found that only seven political–strategic populists, all of whom inherited uniquely enabling conditions, led to democratic breakdown. Similarly, Carrión (Reference Carrión2021) reviewed five Latin American populists, finding that only those who could mobilize supporters and who controlled the state’s repressive apparatus experienced significant backsliding. Representing the ideational side, Ruth-Lovell, Lührmann, and Grahn (Reference Ruth-Lovell, Lührmann and Grahn2019, Reference Ruth-Lovell, Lührmann and Grahn9) found that only a third of populist governments experienced significant backsliding (although none were associated with democratic improvements). Thus, whereas the literature has consistently found that, on average, populist incumbents significantly harm democratic contestation, some are more damaging than others. In this regard, we believe an important control and moderator is time in office. Currently, only a few quantitative researchers have included length of tenure as a control in their models, generally without elaborating on their results (e.g., Huber and Schimpf Reference Huber and Schimpf2017; Ruth-Lovell and Grahn Reference Ruth-Lovell and Grahn2023).

RESULTS: POPULISM AS CORRECTIVE

In contrast, empirical research generally has not confirmed a strong, beneficial effect of populist incumbents on participation and democratic representation. These findings are not necessarily insignificant or negative but rather inconsistent, and the general sense is that more research is needed.

The seminal work was an article by Mudde and Rovira Kaltwasser (Reference Mudde and Kaltwasser2013), which expanded on an earlier argument by Filc (Reference Filc2009) concerning the existence of “inclusive” and “exclusive” populists in Israel. Mudde and Rovira Kaltwasser compared four populist governments in Europe and Latin America to show that left-wing populists are more likely than right-wing populists to bring about improvements in participation and political representation because they tend to defend the rights of excluded minorities.

Although some country studies seem to confirm this argument (e.g., Stavrakakis and Katsambekis Reference Stavrakakis and Katsambekis2014), large-N studies from both approaches have had mixed results. For example, in their political–strategic study of populism’s effects in Latin America, Houle and Kenny (Reference Houle and Kenny2018) failed to find any effect of populism on voter turnout. However, they did not check for the moderating impact of left and right ideology. Similarly, Kenny’s global political–strategic study (2017) found null results regarding populism and voter turnout—but so did Huber and Ruth’s (Reference Huber and Ruth2017) European study, which drew from the ideational approach. In contrast, Leininger and Meijers (Reference Leininger and Meijers2021) found that populism is associated with increased turnout—but only in Central and Eastern Europe—and the effect is not specific to left-wing populists. Furthermore, Juon and Bochsler’s (Reference Juon and Bochsler2020) ideational cross-regional study found that left populists have beneficial effects for participation.

The situation is similar for democratic representation. In their study of Latin America (where left-wing populists predominate), Ruth-Lovell and Hawkins (Reference Ruth-Lovell, Hawkins, Heinisch, Holtz-Bacha and Mazzoleni2021) failed to find any beneficial effects of populism for descriptive and material representation. However, Juon and Bochsler (Reference Juon and Bochsler2020) again found that left populism is positively associated with a broad index of representation. In another cross-regional study, Cole and Schofer (Reference Cole and Schofer2023) found that populist incumbents are associated with policy outcomes that fit their programmatic leanings (i.e., stronger material representation) and that this is true for populists of both the left and the right. All of these studies drew largely from the ideational approach.

Thus, whereas there are tantalizing signs of the positive consequences of populism for democratic participation and representation, these results are much less consistent than for the negative effect on contestation and, overall, there are fewer studies.

OPPORTUNITIES FOR RESEARCH

The populism literature has produced increasingly distinct causal arguments and a strong series of findings on the relationship between populist incumbents and some institutions of liberal democracy (especially contestation). However, scholars have overlooked two opportunities for enhancing research.

The first opportunity is empirical and concerns the use of better measures of populism and larger cross-regional datasets. Quantitative research from both approaches often relies on rough measures of populism that combine measures of rhetoric and organization. This makes it difficult to compare the merits of different theoretical approaches, a problem we elaborate on in this discussion. Furthermore, even recent studies that rely on more careful measures capturing only one of the approaches—for example, Juon and Bochsler’s (Reference Juon and Bochsler2020) ideational measure of populism—tend to rely on expert surveys and, in some instances, assessments by scholars themselves that omit measures of reliability. Finally, almost all of the studies cited in this article use dichotomous measures of populism, and many combine datasets from different regions, where thresholds for “populist” are often inconsistent. We suspect that many of these measures are susceptible to false positives and measurement error, which make it more difficult for quantitative studies to find statistical associations.

Likewise, many studies rely on regional rather than cross-regional datasets. Whereas geographical focus is often a virtue in country-case studies, in quantitative research, this strategy limits variation in key variables, making it more likely to underestimate causal relationships. After all, studies frequently suggest that populism is stronger—both rhetorically and electorally—in developing regions such as Latin America and Central and Eastern Europe (Jenne, Hawkins, and Silva Reference Jenne, Hawkins and Silva2021); that certain ideological versions of populism are sparse in certain regions (Mudde and Rovira Kaltwasser Reference Mudde and Kaltwasser2013); and that the decline in key democratic indicators is overall lower in wealthy, experienced democracies (Papada et al. Reference Papada, Altman, Gastaldi, Kohler, Lundstedt, Natsika, Nord, Sato, Weibrecht and Lindberg2023).

Ultimately, scholars can best eliminate these sources of error with continuous, objective measures across multiple regions. Although we think that direct, objective measures of populism should be the gold standard, they are often difficult and expensive to produce—although they do exist (Jenne, Hawkins, and Silva Reference Jenne, Hawkins and Silva2021). In the meantime, scholars can make greater use of systematic expert surveys that recently have become available, such as V-Party from Varieties of Democracy (Lindberg et al. Reference Lindberg, Düpont, Higashijima, Kavasoglu, Marquardt, Bernhard and Döring2022). Even qualitative researchers can draw from replicable data that provide reliability measures. Doing so not only strengthens our confidence in the association between populism and contestation but also might allow us to discover more consistent associations elsewhere.

A second missed opportunity is scientific. Even after years of populism studies, analysts are often reluctant to directly compare the predictive power of competing theories, especially the ideational and political–strategic approaches. Almost all of the studies cited in this article can connect to one of these two approaches, but no study tests the two approaches together in the same model or by comparing their different causal pathways in a case study. Part of the problem is that only categorical measures of the political–strategic approach exist; any statistical comparisons with continuous ideational measures would be biased toward the ideational approach. Furthermore, as noted previously, several studies use indicators that unsystematically combine different indicators. As a result of these failures, we cannot identify which causal mechanisms are doing the work, which renders results of the literature scientifically ambiguous.