Generation Z: pessimistic and populist? A conjoint experiment on the determinants of populist voting in Spain

Abstract

Generation Z is the most educated and yet pessimistic about the future. At the same time, populist parties have much support among young voters. Do they find an answer to their discouraging socio-economic situation in populist appeals? In this article, we analyse how pessimistic economic expectations shape the preference for populist parties among the young in Spain. By using conjoint experiments, we explore which specific features of populist parties (‘thin’ or ‘thick’ characteristics) are decisive in attracting young and pessimistic voters. Unlike older generations, for whom immigration is the most relevant factor, Generation Z, especially the pessimistic, focuses more on the thin ideological elements of populism. This finding contradicts previous experimental studies, which argue that thin populist characteristics are irrelevant in explaining the general population’s voting behaviour. Ideology plays a significant mediating role, as young pessimists on the left tend to be attracted by anti-elitism, while those on the centre-right by people-centrism.

Introduction

The rise of radical and populist parties in Western democracies and its determinants are among the most pressing challenges for research in political science. There are an increasing number of studies exploring the possible causes of this phenomenon and its potential impact on different contexts and publics (Kriesi and Pappas, 2015; Amengay and Stockemer, 2019; Norris and Inglehart, 2019; Gidron and Hall, 2020; Gomez and Ramiro 2023). Studies focusing on populist radical parties usually distinguish between their ‘thin’ (Mudde, 2004; Stanley, 2008) and ‘thick’ (Pirro et al., 2018; Taggart, 2017) ideological features. While the latter refers to specific positions associated with host ideologies, such as anti-immigration, anti-globalization or pro-redistribution stances, the former is based on the ideas of people-centrism and anti-elitism (Hawkins and Rovira Kaltwasser 2018: 3; Katsambekis, 2022; Zaslove and Meijers, 2024).

In this paper, we explore the role of thin and thick ideological features in determining the choice of voting for a populist party. This question has concerned a recent strand in populism studies seeking causation of the populist vote (Neuner and Wratil, 2022; Castanho Silva et al., 2023; Christensen and Saikkonen, 2022). These authors, using experimental data, have demonstrated the limited effect of populist elements (thin ideology features) of candidates to attract voters. However, it remains unexplored how the effect of these thin features might vary depending on voter type.

Specifically, there is no research considering the extent to which thin and thick ideological features of a populist party or candidate influence the youth vote. We also do not know if poor prospects increase the likelihood of feeling attracted by populism among the youth and how it interacts with the thin/thick division. To fill in these gaps in the literature, we focus on Spain, a country that leads the European ranking in youth unemployment and precariousness and has both right and left populist options on the electoral menu. Do negative socio-economic contexts generate a preference for populist parties especially among the young? Are young people who are pessimistic about their future those who support populist parties the most? Which supply-side drivers (thin or thick) lead them to support populist parties?

This article explores these questions by using a choice-based conjoint design containing thin and thick populist attributes with a sample of 2,003 citizens, including a subsample of 481 respondents belonging to Generation Z (i.e. those aged 18–27). ¹ Our results show that, in general, Generation Z tends to favour parties more due to the thin populist elements than the thick. The thick ideological features, particularly anti-immigration, affect older citizens to a greater degree. The thin elements, in particular, seem to lure younger citizens who are pessimistic about their future. We also show that there are differences among young people according to their ideology, with those on the left being more likely to be attracted by anti-elitism and those on the right by people-centrism.

These findings have implications for the way we study how voters are appealed to by populist discourses. While the literature has predominantly argued that thin features of populism are not as relevant as the thick elements, this paper demonstrates how this appeal might depend on the voter’s profile. In a context of instability and a lack of socio-economic prospects for the new generations, we observe greater support for populist parties among the young, an increase that could be precisely explained by their interest in thin elements such as anti-elitism and people-centrism. Regarding policy implications, this article highlights how unpromising job markets might attract Generation Z to populist options and thus offers a convincing argument for improving youth labour conditions to deter the populist upsurge.

The article is structured as follows. In the following section, we review the current state of the art in populism research and argue for a more careful investigation into the interplay between the supply and demand sides of populism, with special attention to the youth. Next, we discuss the role of individual socio-economic perspectives on the populist vote among Generation Z. The methods section outlines the reasons for the selection of Spain as a case study, the conjoint design, variables and measures used in our analysis. Finally, we discuss the results and conclude by summarising the main implications of this study.

Populist supply and demand: thin or thick ideology?

Three main explanations for the populist party vote have focused on the demand side. First, the economic thesis contends that the socio-economic conditions of individuals are good predictors of the populist vote (Spruyt et al., 2016). It has been argued that populist voters are to be found predominantly among the so-called ‘losers of globalisation’ (Kriesi et al. 2006). Second, the ‘cultural’ thesis, which argues that the rise of post-materialist values, explains the populist vote. For example, in the case of populist radical right parties, Norris and Inglehart (2019) defend the cultural backlash theory, arguing that those voters with anti-immigration attitudes or perceptions against multiculturalism and diversity are more inclined to vote for these parties. Third, populism research has recently explored to what extent the populist attitudes of individuals (e.g. people-centrism or anti-elitism) predict the preference for populist parties (Akkerman et al., 2014; Van Hauwaert and Van Kessel, 2018; Wuttke et al., 2020). Based on these three explanations, survey research has consistently explored the links between demand side characteristics and the populist vote or the populist preference.

Assuming that these theses are partially correct and complementary, a recent advance in populism research seeks to go beyond this approach and concentrates on the relation between the supply features of populist parties and the characteristics of voters (i.e. the demand side). In terms of the supply dimension, populism comprises thin core elements that can be combined with various thick or ‘host’ ideologies (Mudde, 2004; Mudde and Rovira Kaltwasser, 2013; Stanley, 2008). Two thin elements stand out: people-centrism and anti-elitism. The former indicates the extent to which a party defends sovereignty and that even the most important decisions ‘should be an expression of the “volonté Générale” (general will) of the people’ (Mudde, 2007: 23). This strategy of constantly referring to the ‘common people’ and their supposedly shared general interest is a prominent approach employed by populist actors. The latter responds to the Manichean view that society is divided into ‘two homogeneous and antagonistic groups, the pure people versus the corrupt elite’ (Mudde, 2007: 23). This aspect is closely related to people-centrism as it portrays political and economic elites as solely pursuing their own interests and disregarding the well-being of the ‘common people’.

Regarding the thick ideology, the literature identifies three main elements: first, the thin ideology of populism can be combined with positions on economic redistribution, with a subsequent claim for less state and more liberalisation or greater intervention by the state. Second, nativism is an essential component of thick ideologies especially in the case of radical right parties. Nativism refers to the preference for the native-born and their culture in contrast to the foreign-born (Betz, 2019: 112; see also Mudde, 2007: 19). Nativism can be expressed in diffuse ways such as rejecting European integration or globalisation in the defence of the native population (Habersack and Werner, 2023: 867) but always implies a preference for the natives. As recent research shows, this implies that nativists perceive immigrants to be a threat to their country (Kokkonen & Linde, 2022: 4). In its justification, nativism can rely on economic reasons such as welfare chauvinism (Betz, 2019: 117-122) or cultural threats in the last wave of nativism, crystallising especially in anti-Islamic mobilisations (Betz, 2017: 348). Third, in Europe, the thick ideology of populism is usually composed of Euroscepticism, with a diffuse or qualified rejection of European integration or particular EU policy dimensions (Taggart, 2017). These three basic elements of the thick ideology of populism are articulated and expanded depending on the national context in which a political party operates or the ideological leaning on the left or the right of the political spectrum.

In this sense, the aim of this research is to capture the interplay between the supply side—the party’s appeal—and the demand side—voter preferences. However, in ordinary survey research, it is difficult to disentangle the specific features of populist parties that are decisive to attract voters. To comprehensively grasp the mechanisms mediating between voters and populist parties, recent research has relied on experimental evidence. Neuner and Wratil (2022) conducted a conjoint experiment on the determinants of the populist vote and concluded that the thin ideology features (i.e. people-centrism and anti-elitism) of parties do not determine the choice of individuals as much as the thick ideology components of party discourses. Furthermore, those individuals with populist attitudes are no more inclined than others to choose a populist party for its thin ideology, according to the findings of this research. In their replication study, Castanho Silva et al. (2023) reached similar conclusions. Hence, the existent empirical evidence on the determinants of the populist vote suggests that ‘the populist elements’ of a party discourse are not decisive in stimulating a preference for that party, even among individuals with populist attitudes. To put it differently, thick ideological features, such as anti-immigration appeals, nationalism, nativism or economic policy positions, are more determinant than thin-ideology characteristics when citizens opt for a populist party.

Two possible conclusions can be drawn from these findings. On the one hand, it might be argued that populist elements in a political party are secondary to the thick elements in determining voter choice. On the other, a more cautious conclusion would be that the effects of thin populist elements might still be relevant but that they are dependent upon some individual-level characteristics that are yet to be explored. As Christensen and Saikkonen argue, ‘certain actions of populist actors may appeal to segments of the population, while other actions appeal to other segments’ (2022: 2). This article builds on the most recent literature on the role of thin features of populism appealing to a specific segment of the population and in a specific context: namely, the young in a negative socio-economic context.

Populism and young voters

It is well established that young cohorts in Europe, although with some cross-country variance (Kitanova, 2020), vote differently from previous generations (Van der Brug and Kritzinger, 2012; Grasso, 2014). On the one hand, and in line with modernisation theory, the young tend to have more post-materialist values and vote accordingly (Inglehart, 1977; Montero and Torcal, 1994). On the other, they are also more prone to support populist and radical parties than their older counterparts for the following reasons: (i) younger voters have less developed party attachments and tend to experiment more with their vote, supporting new and non-mainstream radical parties to a greater degree than their older counterparts (Arzheimer 2018; Rekker, 2022); (ii) they also have been more exposed to the Great Recession’s impact on the labour market, with their precarious jobs increasing their support for populist radical parties (Zagórski et al., 2021); and (iii) they are less prone to associate these parties with the fading stigma of the authoritarian past (Heyne & Manucci, 2021).

Thus, we argue that younger generations should be less likely to be driven by the thick ideological features of populist radical parties of both left and right. Indeed, they are, broadly speaking, less nativist and less concerned with economic redistribution than their older counterparts (Pollock et al., 2015). The young are also more likely to be attracted by the anti-establishment lure of anti-elitism and people-centrism, given their distrust of elites and protest voting. Although recent evidence (Christensen and Saikkonen, 2022) suggests that young voters are no more attracted to a party’s populist features than the rest of the population, the same study found that candidate profiles that emphasise the will of the people were most popular among Generation Z. In other words, it is precisely among young voters (Generation Z, aged 18–27), who are less integrated into society and more likely to opt for extreme options, rather than among those who are still young but more integrated and with some voting experience (Generation Y, aged 28–40), where thin populist features should be more decisive. Hence, we expect that:

Hypothesis 1: The youngest generation will be more likely to choose a party due to its thin populist features rather than its thick ideological stances.

However, the youth are not a cohesive segment of the population but a highly heterogeneous one due to diverse political preferences. In terms of ideology, while some individual traits (related to socio-economic conditions) may be determinant in explaining the propensity to vote for right-wing populist actors, the same traits might be less powerful or even useless in explaining the vote for left-wing parties (Santana and Rama, 2018). Therefore, it is necessary to consider the specific contexts that define the determinants of the populist vote (see Hawkins and Rovira Kaltwasser, 2018) and the diverse offers among populist parties (Roch, 2021; 2022). For example, left-wing populism is supported to a greater degree by the youngest cohort according to several studies (Foa and Mounk, 2019; Vidal, 2018; Lobera and Roch, 2022), whereas in the case of right-wing populist parties, the effect seems to be context-dependent (Stockemer et al., 2018: 577; Zagórski et al., 2021). Thus, we take into consideration the ideological leanings of the youngest and the context of the emergence of different populist parties.

On the one hand, when analysing the characteristics of left-wing populism, usually a combination of populist thick and thin elements can be observed (Rooduijn and Akkerman, 2017). The predominant issue is economic policy and pro-redistribution positions, combined with a clear anti-elitist element that refers not only to the ‘political’ but also to the ‘economic’ elites (Hameleers et al., 2021). On the supply side, left-wing populists have been found to be much less people-centric than right-wing populists (March 2017). Consequently, we hypothesise that:

Hypothesis 2a: Members of the youngest generation located on the left of the political spectrum will be more likely to choose a party with anti-elite features.

On the other hand, right-wing populism also shows a combination of thick and thin ideological elements. The main thick dimension of these parties is composed of nativism, anti-immigration and, in many cases, Euroscepticism (Mudde, 2007; Rooduijn, 2015; Santana et al., 2020). However, they connect these positions with populist features, with a predominance of people-centric discourse (March, 2017) from a nativist-exclusionary point of view (Marcos-Marne et al., 2021). Thus, for radical right populists, ‘the people’ are closely linked with the concept of ‘nation’ (Mudde, 2007). At the right side of the political spectrum, the ‘people’ tend to be combined with exclusionary identity markers seeking to strengthen the people as a nation rather than the people as plebs, more politicised by left populism (Vergara, 2020; Mudde, 2021; Vulović and Palonen, 2023; Roch and Cordero, 2023). Therefore, we hypothesise that:

Hypothesis 2b: Members of the youngest generations located on the right of the political spectrum will be more likely to choose a party with people-centric features.

Economic expectations and populism

Since at least WWII, a shared idea among capitalist societies searching for progress and enrichment was that new generations should lead a better life than the previous, with individuals expected to meet or exceed the living standards of their parents’ generation (Hyman, 1942). The Great Recession has challenged this expectation, with an increasingly shared view among young people that they will not be able to achieve their goals in their lifetime due to increasing inequality and precariousness, particularly in the youth labour market. But could this lack of prospects trigger a preference for populist parties?

There is conflicting evidence on the extent to which an individual’s economic expectations influence the likelihood of voting for a populist party. While Berman (2021: 75) argues that explanations based on economic grievances are not good predictors of voting for far-right populist parties, Spruyt et al. (2016) show that the socially marginalised are more likely to vote for populist options. Similarly, Gidron and Hall (2020) also show that those who feel socially marginalised (though not necessarily with objectively low economic status) are more likely to vote for a populist party. Indeed, the accumulating evidence seems to point to a recent trend: objective socio-economic characteristics are losing ground as an explanatory factor in populism studies in favour of more subjective and intersubjective measures of how voters feel about their economic situation and their hopes for the future (Gidron and Hall, 2020; Oesch and Vigna, 2023).

These subjective measures have been operationalised in different ways, such as feelings of lack of recognition (Steiner et al., 2023) or lack of control (Heinisch and Jansesberger, 2022). Recent research has focused on determinants such as pessimism about the future and negative perceptions of one’s own socio-economic status (Gidron and Hall, 2020) or perceived ‘nostalgic deprivation’ and ‘status threat’ (Gest et al., 2018). At the same time, recent studies highlight that the triggering mechanism that activates party choice or political preferences is more likely to be related to a perceived (lack of) intergenerational mobility than to immediate material interests associated with working conditions (Evans et al., 2022; Langsæther et al., 2022). In fact, Bolet (2023) finds that subjective social decline explains support for both radical right- and left-wing populist parties.

Thus, anxiety provoked by a lack of ‘optimistic prospects’ is said to be at the root of support for all (populist) radical parties, with the relationship with radical right voting being a more studied phenomenon. Status comparisons or discordance are linked to relative deprivation theories with the benchmarks for comparison being either group or individual status in the past, that is, 30 years ago (Gidron and Hall, 2017; Gest et al., 2018), childhood expectations (Kurer and Van Staalduinen, 2022) or the status of parents (Bolet, 2023). This status anxiety has four important characteristics for explaining support for populist radical parties. First, it is not only of an economic nature, as it entails such political and social terms as efficacy and prestige and thus bridges the existing economic and cultural explanations of the populist vote. Second, it lies in the realm of the subjective, and, thus, it does not have to be objectively real, at the same time incorporating the emotional underpinnings of populist voting. Third, it can be operationalised as expectation regarding the future and not only as the experience (and the former have been found to have stronger effects; see Im et al., 2023). Fourth, it is ideologically neutral, and as such, it applies to explanations of both the radical left and radical right support.

In conclusion, given the lack of expectations and the slowdown in intergenerational mobility compared to previous generations, we expect those young voters who are most pessimistic about their economic prospects to lean more towards the thin populist characteristics of a party (people-centrism and anti-elitism). Consequently, we expect that:

Hypothesis 3: Members of the younger generations who are pessimistic about their economic prospects will be more likely to choose parties with thin populist characteristics.

Method, data, case of study, design and operationalisation

Method

To test these hypotheses, we use a survey experimental design based on choice-based conjoint analysis (see Raghavarao et al., 2010). The rationale for using a conjoint design lies in the limitation of ordinary survey experiments to analyse items with multiple dimensions (Hainmueller et al., 2014). The latter can capture the treatment effect as a whole but cannot identify which specific component is driving the effect of the manipulation (Hainmueller et al., 2014). This is a major drawback when it comes to disentangling the specific features of populist parties that are crucial for attracting voters. In a choice-based conjoint design, participants are forced to choose between two options with different attributes that vary randomly. In this way, we can identify the effect of variation in each attribute on the probability of choosing one or the other alternative simultaneously, that is, the marginal effect of each element (see Hainmueller et al., 2014). For this reason, conjoint analysis has been widely used in recent years to identify the mechanisms that mediate between voters and populist parties (e.g. Ferrari 2024; Castanho Silva et al., 2023; Christensen & Saikkonen, 2022; Neuner & Wratil, 2022). Following this novel line in the literature, we combine multiple thin and thick populist attributes and ask respondents to indicate which of the two party’s profiles they would choose to vote for in a hypothetical election.

Data

The experiment was embedded in an online survey fielded by the polling company Netquest in July 2022 in Spain on a representative sample of the Spanish population in terms of gender, age, size of the place of residence, region (Comunidad Autónoma), level of education and social class (N = 2,003). This sample includes a subset of 481 respondents aged between 18 and 27 (Generation Z) and a subset of 550 between 28 and 40 (Generation Y). Despite the wide variety of classifications, we base our categorisation on that proposed by the Spanish public institution ‘National Youth Institute’ (Cerezo, 2016), a classification specifically designed for the Spanish case. According to this institution, Generation Z are considered those born from 1995 onwards and Generation Y those born between 1980 and 1995. We slightly adapt it to our sample, considering Generation Z as those born from 1996 onwards and Generation Y as those born between 1982 and 1995.

Case of study

Extant similar studies have focused on the USA (Ferrari 2024; Castanho Silva et al., 2023) and Germany (Neuner and Wratil, 2022). Thus, on top of testing up to which point their findings apply also to the Southern European context, the reason for studying the case of Spain lies in its appropriate characteristics on both the supply and demand sides to test our hypotheses. First, on the demand side, economic insecurity, relative deprivation and anxiety due to a lack of optimistic economic prospects are said to be one of the main determinants of voting for populist parties (Gidron and Hall, 2020; Kurer and Van Staalduinen, 2022). Spanish young people from both Generations Y and Z have seen their expectations for the future evaporate in the face of a succession of economic crises, from the Great Recession to the Covid-19 pandemic. In fact, Spain is currently one of the countries with the highest rate of young people with higher education but also the highest youth unemployment (Eurostat, 2024). As a result, young Spaniards tend to be pessimistic about their future due to a lack of opportunities and intergenerational mobility (Marqués Perales and Gil-Hernández, 2024). As a consequence, the context in which Generations Y and Z were socialised gave rise to a double-sided situation in relation to youth: it entailed the acquisition of political knowledge for many young people (e.g. Castañeda, 2012); and, as this politicisation was linked to a populist discourse, the younger generations in Spain were those who may have incorporated populism into their political attitudes and voting behaviour to a greater degree (Galais, 2014).

Like in other southern European countries, the Eurozone crisis meant massive job losses, shrinking GDP and disappearing hopes for many young people (Antentas, 2015). Spain was one of the countries hardest hit by the crisis and is an emblematic case of youth precarity. For Generation Y, the high youth unemployment rates during the Great Recession had far-reaching social and economic consequences during their first years in the labour market, with many young Spaniards facing challenges in finding stable employment, launching their careers and achieving financial independence. The poor labour market conditions and the economy were the context in which Generation Z was socialised, having from a very early age a pessimistic vision of the opportunities that they would face after their educational period.

The consequences of the Great Recession prompted the emergence of social movements with strong populist components—anti-elitism and people-centrism (Aslanidis, 2016)—such as the 15M and Indignados movements in 2011, which challenged the two-party system and the political system itself (Sampedro and Lobera, 2014). Populist traits can be clearly observed in the manifestos of some of the largest organisations of the 15M movement, such as ‘Democracia Real Ya’ (Real Democracy Now) (Hughes, 2011). Although Generation Z was not socialised around these events, the Great Recession, as in other European countries, had a major and long-lasting impact on the way the following generations of Spaniards engaged with social movements and politics (García-Albacete et al., 2015; Benedicto and Ramos, 2018).

Second, on the supply side, the party system has mutated in the last decade from a two-party system to a multi-party system with a left-wing populist party, Podemos, and another right-wing populist party, VOX (Bosch and Durán, 2019; Hughes, 2011; Marcos-Marne et al., 2021). Podemos, a party that emerged from the Great Recession, the 15M movement and the Indignados movement, was formed in 2014 with a strategy that included both pro-redistribution and anti-elitism (Kioupkiolis, 2016; Ramiro and Gomez, 2017). Podemos’ strategy was strongly linked to populist appeals such as the use of the word ‘casta’ (caste) to describe the political and economic elites, the slogan ‘no nos representan’ (they don’t represent us) and criticism of the role played by elites during the transition to democracy (Bosch and Duran, 2019). Moreover, although Podemos’ populist stance has diminished since it joined the governing coalition in 2019, it is still alive with the new designation of ‘Sumar’ (add up) and prominent in the party leaders’ discourse on the judicial and media elites (Schwörer, 2022).

Existing evidence shows that young Spaniards are clearly more likely to support the populist radical left Podemos but also the populist radical right Vox (Turnbull Dugarte et al., 2020; Rama et al., 2021b; Heyne & Manucci, 2021) compared to their older counterparts. Both Podemos and Vox emerged in the aftermath of the Great Recession. Although both utilised an anti-elite and people-centric discourse, in the case of Vox, the discourse was characterised by nativism, which led the party to advocate for a participatory democracy of ‘the Spanish people’. Just like other populist radical right parties in Europe, Vox is considered a nativist and populist party (Mudde 2007; Turnbull Dugarte et al., 2020), with frequent appeals to people-centric messages (Rama et al., 2021a). In this sense, Vox’s discourse is not deployed to criticise the elite as a homogeneous group, as is the case with Podemos. The critique of radical right-wing populism in Spain focuses on more specific elites, particularly those on the left and those associated with regional nationalism. Like most radical right parties in Europe (see Stavrakakis et al., 2017; Roch, 2021), Vox’s anti-elitism is less intense and peripheral in its overall discourse, instead choosing to focus its anger on scapegoats such as illegal migrants, leftists, LGBTI individuals or other groups excluded by the ‘legitimate Spanish people’. However, this limited anti-elitism does not prevent the party Vox from resorting to people-centrism by referring to ‘the people’ or ‘the Spaniards’ in a much more widespread manner in its speeches (Marcos-Marne et al., 2021).

Thus, while Generation Y grew up in a period of economic prosperity with a stable bipartisan system and high levels of political satisfaction, their entry into the labour market was disrupted by a global crisis of enormous depth, triggering an unprecedented political shift in which populism started playing a crucial role. On the other hand, Generation Z was already socialised in an unfavourable economic context, marked by successive economic crises. In the political arena, this generation was socialised in a context where the presence of populist political options was already a reality and where the relation with politics was more cynical from the beginning. This article examines to what extent these generations, particularly Generation Z, are attracted to populist elements of the discourse in an environment of low economic expectations.

Conjoint design

With regard to the design of the conjoint experiment, participants were required in four consecutive rounds to choose between two political parties (A and B, without logo or name) with seven different attributes: two thin populist characteristics (people-centrism and anti-elitism), three thick ideological characteristics (economic interventionism, position on immigration, Euroscepticism) together with support for liberal democracy (in the civil liberties dimension, measured as support for minority rights vs. majoritarianism) and regional autonomy, presented in random order to each respondent (but in the same order in all four rounds). The levels of each attribute were also randomised, with no restrictions on the possible profiles. In other words, the conjoint has a fully randomised design (see Hainmueller et al., 2014). As in all choice-based conjoint analyses, the dependent variable is the choice outcome, in this case, party choice from the two possible alternatives (the exact wording is ‘If you had to choose between these two parties, which one would you choose, A or B?’), and the independent variables are the attributes of each party. All attributes have two levels, as shown in Table 1.

Table 1. Conjoint experiment design

Operationalisation

On the supply side, conjoint attributes are divided into thin and thick populist ideological characteristics (Mudde, 2007). Regarding the former, people-centrism indicates the extent to which a party defends popular sovereignty and that even the most important decisions ‘should be an expression of the “volonté Générale” (general will) of the people’ (Mudde, 2007: 23). We have operationalised this by asking whether important decisions should be taken by politicians or by the people. Anti-elitism responds to the Manichean view that society is divided into ‘two homogeneous and antagonistic groups, the pure people versus the corrupt elite’ (Mudde, 2007: 23). Therefore, we have asked whether the elite, both economic and political, follow their own interests, as opposed to those of the people, or seek the common interest.

As for the latter, while positions on economic interventionism, immigration and European integration are the core components of thick ideological features of populist parties throughout Europe, ‘stance on regional autonomy’ is an essential element of the Spanish electoral competition. Nativism is measured with a single indicator on the position regarding immigration (being economic or culturally mediated), in line with most recent research (Kokkonen & Linde, 2022; Habersack and Werner, 2023). The left/right ideological scale was left out deliberately so that partisanship does not minimise the effect of the other factors and bias the outcome choice (see Kirkland & Coppock, 2018). Support for majoritarianism (as an alternative to liberal democracy) is also accounted for, as individuals with stronger populist attitudes have been found to be highly supportive of forms of unconstrained majoritarian rule (Zaslove and Meijers, 2024) and turn a blind eye to violations of key democratic principles and defying minority rights (Svolik et al., 2023).

On the demand side, we assess the effects of party attributes on the ‘forced choice’ separately by age, divided into three categories: Generation Z (18–27 years old), Generation Y (28–40) and older individuals (41+). We also consider the perceived socio-economic expectations relative to their parents’ situation to analyse the extent to which this pessimism or optimism determines their party choice. ² We recoded the original 5-point scale into two subgroups: optimists (those who think they will do ‘much better’, ‘better’ or ‘same’ than their parents) and pessimists (those who think they will do ‘somewhat worse’ or ‘much worse’). Lastly, we compare the effects according to ideological self-placement: left (0–2), centre-left (3–4), centre (5), centre-right (6–7) and right (8–10), on the traditionally employed left-right ideology self-placement, using an 11-point scale.

Results

To present the results of the choice-based conjoint analysis in the main text, we estimate the average marginal component effects (AMCE). The AMCE represents the marginal effect of a given attribute of the conjoint averaged over the joint distribution of the remaining attributes (Hainmueller et al., 2014). Errors are clustered at the level of the respondent, as the choice outcomes of the four party pairs to which the respondent had to respond are not independent for the same participant. To disentangle the mechanisms between supply and demand in populist voting and to test our hypotheses, we include subset analyses between conjoint attributes and voter characteristics and political preferences. Hence, as a robustness check, following the recommendation of Leeper et al. (2020) on the examination of differences between subgroups, we replicate the analyses using marginal means (MMs) instead of AMCE, which confirm the key results (see Online Appendix).

First, Figure 1 shows the AMCEs and 95% confidence intervals for each party attribute for all respondents and by subgroup according to their generation (Figure A1 in the Online Appendix replicates these models using MMs). At first glance, the results seem to suggest that the youngest (Generation Z) are more attracted to thin aspects of populist ideology than to the thick ideological components. Partially confirming Hypothesis 1, Generation Z is more likely to vote for a people-centric party than for one that advocates that important decisions should be made by politicians (confirmed using both AMCEs and MMs). For this generation, the average probability of voting for a party is 0.05 higher at the 99% confidence level (standard error of 0.02) if it argues that important decisions should be taken by citizens. Notably, this effect is significant only for the youngest generation, as for previous generations it is negative and non-significant. The remaining attributes—notice in particular the thick populist features—have no significant effect on Generation Z. Therefore, Hypothesis 1 is only partially confirmed, given that anti-elitism has no significant average effect among the youngest citizens. As for the other age subgroups, while no attribute has a significant coefficient for Generation Y, we find that for Generation 41+ the average probability of voting for a party that thinks the effects of immigration are negative rather than positive is 0.03 higher, with a standard error of 0.01. Thus, older Spaniards seem to prefer a party with a negative view of immigration, a thick component. This translates into the nativist component being the only significant predictor of party support for the whole sample.

Figure 1. Average marginal component effect of party attributes, by generation.

Regarding the second hypothesis, Figure 2 focuses on the youngest according to their ideology (see Figure A2 in the Online Appendix replicating the analysis using MMs). First, while before, on average, people-centric parties attracted Generation Z more than non-people-centric parties, once ideology is included, this effect is only significant for the subgroup of centre-right young people. Thus, Hypothesis 2b is fulfilled, although again only partially, as the effect is not significant for those at the extreme right of the ideological scale (both using AMCEs and MMs). The average probability of voting for a party increases by more than 12% for centre-right members of Generation Z when including the people-centrism element compared to not. In terms of the thick ideological components, centre-right members of Generation Z seem to prefer on average parties with liberal economic positions by 0.09 more than parties that defend higher taxes and expanded public services. Second, with respect to young leftists, we again find that Hypothesis 2a is partially satisfied: while young people at the left end of the scale are on average almost 9% more likely to vote for parties with anti-elitist discourses, this is not the case for young people in the centre-left subgroup, where the coefficient is non-significant (both using MMs and AMCEs).

Figure 2. Average marginal component effects of party attributes for Generation Z, by ideology.

While the latter finding is intuitive and falls in line with our argument of anti-elitism being salient for the leftist subgroups of the youth (also keeping in mind that the distribution on the L-R ideological self-placement scale is skewed to the left in Spain with those locating themselves on the centre-left forming a more heterogeneous group), the result of the centre-right and not the right youth being more prone to be lured by people-centric attitudes is puzzling. One supposition would be that it is connected with the type of populist appeal expressed by the operationalisation of ‘the people’. Results seem to indicate that radical right voters are not inclined to favour the people when making the most important decisions about politics. In the context of Spanish politics, this claim resonates with the previous politicisation by left populism and would have prevented radical right voters to identify with this claim. As recent scholarship on the radical right in Spain shows (Marcos-Marne et al., 2021; Ramos-González and Ortiz, 2022), but also more generally in Europe (Mudde, 2021), nationalism is the core element of these electorates. Thus, since people-centrism is not clearly connotated with the central role of the nation, it can be perceived as democratic expansion and democratisation of collective decision-making processes, a classic demand of radical left and left populism in Spain (Roch, 2021), which might be one of the explanations of the less significant attachment to the specific operationalisation of people-centrism among the radical right in this study. These findings put some constraint on the generalisation of our study as they seem to be context-dependent.

In any case, the fact that both thin populist sub-dimensions—people-centrism (for those on the centre-right) and anti-elitism (for those on the left)—have a significant effect, and that immigration, European integration and centralisation are less relevant issues for members of Generation Z, regardless of their ideological leanings, reinforces Hypothesis 1.

For previous generations, most attributes do not seem to determine the likelihood of voting for a party when individuals are grouped according to their ideology (Figures A7 and A9 using AMCEs and Figures A8 and A10 using MMs). It is only when assessing the party preferences of Generation Y (Figures A7 and A8) that we find that the effect of anti-elitism differs from that of Generation Z: in this case, far-right members are more likely to choose a party with an anti-elitist discourse, while for all other ideological positions, the effect is not significant. This seems to indicate that anti-elitist parties have a stronger mobilising effect among younger far-left voters and among middle-aged far-right voters. Future research should further investigate the role of the supply side in this phenomenon.

Moving on, as argued above, when studying the populist vote of young people, it is necessary to consider their expectations for the future. Perceptions of relative deprivation and lack of future prospects are essential for understanding the interplay between supply and demand for young people. For this reason, Figure 3 disentangles the effects of party attributes by age and optimism about the future. The previously found positive effect of people-centrism among members of Generation Z is now only significant for those young people who believe that they will never achieve the same status as their parents: among ‘young pessimists’, the average probability of voting for these parties increases by 0.05 at the 95% confidence level.

Figure 3. Average marginal component effects of party attributes for Generation Z, by economic expectations.

In Figure 4 (and Figure A4 in the Online Appendix), we have also added ideology as a grouping variable (on top of economic expectations), as we have seen that it is relevant for disentangling the effects of the thin populist sub-dimensions from the thick elements. The results show that an anti-elite stance has a significant effect of more than 0.12 on the probability of voting for a party for those in Generation Z who are on the extreme left and who believe that they will never achieve the same social position as their parents. Moreover, among the youngest, those who are pessimistic and right-wing are more likely to vote for parties that do not support the discourse of territorial (re)centralisation in Spain. Thus, although not all attributes are always significant, the results suggest that young pessimists are more attracted to thin populist sub-dimensions (Hypothesis 3): either people-centrism (across the ideological spectrum) or anti-elitism (for those on the left).

Figure 4. Average marginal component effects of party attributes for Generation Z with pessimistic economic expectations, by ideology.

Conclusions

The rise of populist parties and the electoral behaviour of young people, separately, have been extensively discussed in the extant literature. However, only a handful of studies have assessed support for populist parties among the young, and we still do not know which features of populist parties attract the youth nor up to which point they depend on their ideology or socio-economic situation. This article fills this gap by adding a new perspective to recent studies that use experimental approaches to understand the interplay between supply and demand in populist voting.

The focus is on Spanish youth, particularly on the youngest voters in Generation Z, who have especially suffered economic distress in the last decade. Existing similar studies were conducted in the USA (Ferrari 2024; Castanho Silva et al., 2023) and Germany (Neuner and Wratil, 2022), so testing the effects of thin/thick features of the populist supply in the Spanish context is valuable per se. Although the study focuses on data from Spain, its findings shed light on a possible broader trend. Younger generations in Western countries, especially those in Southern European nations, have lived through the Great Recession and have felt its political repercussions firsthand. However, given some idiosyncrasies and inconsistencies with the findings of existing studies, we found that replicating this study in other contexts would be beneficial to see whether our findings are conditioned by the particular socio-economic context of Spain, or Southern Europe more broadly, or whether they also hold in other parts of Europe.

Our contribution to the understanding of populist appeals is threefold, as we offer an assessment of the role of (1) the thin/thick features among the youth, (2) the thin/thick features among the youth by their ideology and (3) the thin/thick features among the youth by their ideology and the level of pessimism of socio-economic expectations. In line with the previous literature on intergenerational differences in vote choice, we find that younger cohorts exhibit differentiated behaviour with respect to populist and ideological supply characteristics. What we show is that the thin populist features, while irrelevant for vote choice among older people, are crucial for the youth and that these effects vary by their ideology and level of pessimism.

In a conjoint experiment analysis, we demonstrate that for the young (Generation Z), issues such as immigration, EU integration or protection of minority rights have less impact on the decision of which party to vote for. Conversely, the thin ideological key sub-dimensions of populism—people-centrism and anti-elitism—are the strongest determinants for party choice among the young. On the one hand, Generation Z, especially if they are centre-right and think they will not achieve the socio-economic status of their parents, are more likely to choose parties with people-centric stances. On the other hand, anti-elitism appears to be a significant predictor of party choice among left-leaning members of Generation Z, especially among those who are pessimistic about their economic prospects.

Regarding the thick ideological party features, we find that immigration is only salient for the older generations (particularly for those on the far right), while economic liberalism and (re)centralisation seem to influence the choice of the youngest (Generation Z) located on the centre-right.

In sum, our results provide a completely different perspective from other experimental approaches that studied the role of the thin and thick components of populist parties in attracting voters. While previous studies found that it is not the thin ideological dimension (i.e. people-centrism, anti-elitism) of the parties that determines individual choices, but rather the thick ideological components of the party discourse, such as anti-immigration concerns (i.e. Castanho Silva et al., 2023; Neuner & Wratil, 2022), our research shows that this is not the case when the effect is considered separately across different segments of society (see Christensen & Saikkonen, 2022). The results show that, in general, Generation Z tends to base their choice of preferred party on thin populist elements rather than thick, especially among pessimists and those who base their preferences on people-centric party features.

The study is subject to a number of limitations, both theoretical and methodological, resulting from the chosen research design. On the one hand, with regard to the conceptualisation and operationalisation of the concept of populism, our paper engages with other studies within the literature on populism as ideology (i.e. Mudde, 2007), which has become the dominant approach in political science in recent decades. However, there are other currents in the study of populism that this paper does not address (see Moffitt & Tormey, 2014). Our design is based on the most recent research within the ideation approach that suggests a minimalist conception of the thin ideology (with people-centrism and anti-elitism as its main elements) (Zaslove and Meijers, 2024; De la Peña, 2023). Nevertheless, alternative models can be tested by including anti-pluralism or Manichean outlook as other core elements of the thin ideology in future research.

Furthermore, the dimension of anti-politics as an essential feature of populism seems to gain traction in the field of populism studies (Taggart, 2018; Robinson, 2023), and it is worth being explored in future research. With regard to the thick ideology of populism, we have selected three elements that the literature has repeatedly identified as essential within these parties and which also allow us to identify the parties within the ideological spectrum: economic redistribution, nativism and Euroscepticism. However, it remains for future research to study other dimensions that have also been identified with populism, such as authoritarianism, especially in relation to the defence of traditional values (i.e. LGBTI issues). Also, the role of populist attitudes determining the vote for populist parties has been assessed elsewhere (Castanho Silva et al., 2023; Neuner & Wratil, 2022), but it has not been studied for particular segments of the population (especially the youth) yet.

Moreover, as we advanced above in the explanation of the puzzling effect of centre-right and not radical right youth being more prone to choose a people-centric party, different operationalisations of people-centrism can be also tested in future research to evaluate the most effective ways to capture the notion of ‘the people’ for radical right voters, probably emphasising the central role of the nation rather than people as plebs. More studies are needed to understand how general the findings of this study are and whether they depend more on the study design or the particularities of the context.

On the other hand, in methodological terms, although the experimental design allowed us to disentangle the marginal effect of each populist trait, it also has limitations. Given the cross-sectional character of the study, generational or life-cycle effects cannot be distinguished. Future longitudinal studies may address this issue. Additionally, future research should test the external validity of our results: whether the relationships we find between supply and demand, especially the differential effect with respect to age, ideology and economic prospects, hold in a voter survey in a real election context. Similarly, from the supply side, a future line of study could be to investigate whether populist parties adopt different strategies or emphasise their thin or thick elements depending on whether they are targeting one or another segment of the population.

In conclusion, our results underline the need to pay more attention to the heterogeneous effects of populist sub-dimensions by generations, together with the mediating role of ideology and economic expectations. In particular, we show that for younger generations, the thick dimension of populism is not the most important issue in attracting their support, meaning that support is driven more by the way these parties use the thin sub-dimensions of populism than by the specific policies they advocate. Whether these cohorts will continue to be lured by purely populist appeals in the future or change their preferences due to life-cycle effects is an open question.