The Five Star Movement

Abstract

The success of the Italian party Five Star Movement (M5S) has been broadly attributed to its ability to occupy the space of radical protest against “old politics”. Due to the party’s criticism, its charismatic leadership, and its aggressive electoral campaigns, the M5S has been labeled as a populist. The unexpected result of 2013 election raises crucial theoretical questions: To what extent does the M5S electorate reflect the characteristics of a protest vote? To what extent was it also a vote driven by values, by individual evaluations on a specific political issue? The first part of the article aims to investigate the extent of negative political feelings among M5S’ voters. To disentangle the meaning and impact of protest, we distinguish two dimensions: the “system discontent” and the “élite discontent,” referring to both general and focalized images, sentiments toward and the representation of political institutions, voter power, and government performances. In the second part, we bring to the analysis a further explanation based on the theory of issue voting. The goal is to measure whether voters have chosen M5S purely because of their political resentment or also given that they shared a similar position on a number of crucial policies emphasized in the electoral campaign.

Keywords

political parties protest voting issue voting populism Five Star Movement

Introduction

A new party is on the scene in Italian and European politics: the Five Star Movement (M5S). But what kind of party is it precisely? Since 2009, the M5S has enjoyed increasing electoral success, culminating in its first position in the 2013 Italian general elections (25.6%), and an impressive showing at the 2014 European elections (21.2%). The M5S electoral success has often been attributed to its ability to occupy the space of radical protest against the forces of the “old politics,” identified primarily with the existing parties and their leaders (Corbetta and Gualmini, 2013). The combined effect of this antiparty stance, a (quasi) charismatic leadership and an aggressive electoral campaign have contributed to a labeling of the M5S (perhaps too simply) as a populist party. The populist tradition has a long history in Italy (Passarelli, 2016; Tarchi, 2003), and the recent decline of those parties that have recently exploited this “political capital” (Northern League, People of Freedoms) (Albertazzi and McDonnell, 2015; Passarelli and Tuorto, 2012; Passarelli and Tuorto, 2013; Raniolo, 2006) created a gap for such a proposal. Although some populist traits are present in the movement’s rhetoric and political action, such as the dualism of “we” (the people) against “them” (the politicians), which basically represents the core of populist thinking (Canovan, 1981; Hermet, 2001), the reference to populism in the case of M5S might be considered somewhat misleading. The label seems too vague and too general, if not impressionistic and not measurable. The axiom that “this is a populist party, so its voters are populist as well” is of no help if the aim is to understand what underlies electoral choices. Populism is often treated as a catch-all category including parties and voters alike. In this sense, the attention is focused mainly on parties’ characteristics and strategies rather than on voters’ motivations. On the contrary, the concept of protest voting, which has also been widely used to identify Beppe Grillo’s supporters, would seem to be more empirically grounded and useful when the attempt is to explain voting behavior. Since the beginning, the M5S has deliberately focused its campaign and identity on the “protest” (Biorcio and Natale 2013; Bordignon and Ceccarini, 2013; Itanes, 2013), and protest has worked as a key driver for M5S voters. Several aspects characterize political parties: leadership, organization, ideology, or political culture strategies. This article deliberately deals with only one aspect: the voting behavior. Focusing on protest, the attention falls on concepts such as political distrust, discontent, resentment, and all those attitudes and opinions that citizens develop and address to the democratic system, the “establishment,” the élite, and the (incumbent) governments or, in general, to the “old” parties. Did antipolitical sentiment drive the M5S voters? Was their electoral choice mainly a protest vote or a vote driven by specific policy preferences?

A definition of the protest vote

In order to measure the effectiveness of the “protest” component in voter behavior, it is crucial to clarify and define what political protest is, against which objects or actors it is addressed and which voters make use of such arguments and resources to orient themselves in the electoral arena.

The political protest is generally identified with unconventional, anomic, or nonstandard mobilization (Almond and Powell 1966; Barnes et al. 1979). According to classical literature, this activity involves those who lack access to the resources of organized pressure groups or those whose values conflict with positions expressed by the dominant elite. The concept of protest refers to both behaviors and attitudes or feelings. Through demonstrations, petitions, and boycotting, people express their grievances or discontent (Opp 2009) but also their need to influence the political system, using traditional or innovative repertoires of action, outside institutional channels, or challenging the existing political system from the inside. The object of the protest can be the regime, the political community, the system, the policies, and the political parties. Moreover, reconsidering the concept of political support theorized by David Easton, scholars have found a relationship between antisystem orientations, political dissatisfaction, and protest action, even though empirical evidences indicate that not all the countries follow this direction.

Once the analytical dimensions have been laid out, it becomes fruitful to attempt to test the hypothesis of the protest vote in the case of the M5S. From a Downsian perspective, protest voting is basically irrational. If voters do not cast their votes for the party that is closest to their policy preferences, then they should not vote at all. On the other hand, if voters no longer prefer party A, but instead rank party B higher, then protest voting disappears in favor of a simple switch of alignments (Kang, 2004). Political protest in disaffected voters may produce both deliberate abstention and voting for an “extremist” party or a new or outcast party. Politically dissatisfied voters may temporarily withdraw their support from their preferred candidate/party even if that party has a good chance of winning, in the hope that this signal of disaffection will lead to downstream improvements in that candidate’s/party’s performance (Kselman and Niou, 2011). This motive makes protest voting an expression of voice rather than an exit, a strategic behavior known under the term “voting with the boot” (Van der Eijk and Franklin, 1996). Moreover, protest voting could reflect not only and simply an expression of disaffection, but it should also include an element of attraction for “different” parties. But which type of parties? First and foremost, the antisystem parties. According to Sartori (1976: 133), a party qualifies as antisystem if it “would not change—if it could—the government, but the very system of government.” As a further specification, Sartori adds that “not all the anti-system parties are such in the same sense: the negation covers, or may cover, a wide span of different attitudes ranging from ‘alienation’ and total refusal to ‘protest’” (Sartori, 1976: 117). This definition applies above all to extreme leftist and rightist parties, but secessionist and religious fundamentalist parties can also be included in the category and, more generally, parties that undermine the legitimacy of the regime in which they operate, like populist or new radical right parties (Ignazi, 2003; Mudde, 2007; van Kessel, 2015).

Due to the political and conceptual overlapping between protest voting and political disaffection or distrust, a conceptual clarification is necessary when the protest vote is considered in relation to, and sometimes as a synonym of, apathy. As reported in other studies (De Luca, 1995; Kang, 2004), protest voting is not in itself an apathetic behavior; it assumes the configuration of a complete electoral act performed as an expression of disaffection against the established parties and/or those for who they usually voted. While political apathy refers to (political) indifference and a sense of estrangement from (or rejection of) the prevailing political system, political protest seems to be more strictly linked to a condition of political alienation and implies a certain (minimal) connection with the political system, even when it is configured as alienation against all.

The debate on the protest vote is flourishing. There is no unique and agreed definition of such a phenomenon nor is there an accepted corresponding list of parties that should attract protest. Nonetheless, certain common features and methodological tools may be used to measure it. In literature, the “protest vote” is generally considered a vote primarily cast to scare the élite but also a vote in which political attitudes are expected to be of minor importance (Van der Brug et al., 2000). The idea of protest voting necessitates something against which voters are protesting, some kind of disapproval, the result of political distrust (Bergh, 2004). A vote for nonestablished parties incorporates in its antielitist position a certain amount of dissatisfaction, and it is expected to frighten or shock the élites from mainstream parties (Ignazi, 1992; 1996; Pedersen, 1982; Poguntke, 1987; Van Biezen, 2003; Van der Brug and Fennema, 2007). This was particularly true since the 2008 economic crisis, when the lack of responsiveness of established parties has provided a chance to populist discourse and the emergence of new political actors (Kriesi 2012; Kriesi and Pappas 2015). This scenario fits well with Italy, whose governments, under the European Union (EU) pressure, have adopted economic reforms and privatization strategies, with the results of high degree of popular discontent, electoral volatility, and punishment of both incumbent and opposition (Passarelli and Tuorto, 2014; Vegetti et al., 2013).

The definition of protest vote allows us to narrow down the field of interest to a vote (a) against the establishment and (b) not driven by policy preferences. With regard to the first aspect (vote against), a protest in the ballot box is given when voters make a choice expressing a certain amount of radicalism, criticism, and dissatisfaction against the established parties in the national parliament. Those feelings can be expressed even without any sentiment of true and genuine sympathy for a party that they could vote for (Bowler and Lanoue, 1992; Van der Brug et al., 2000; Van der Eijk and Franklin, 2009). Protest voting may be defined as the level of dissatisfaction with how the established parties and system, as well as the government and the parliament, have performed. Based on these different objects of protest, Bergh (2004) has differentiated political protest into two distinct but interrelated dimensions: the “system discontent” and the “élite discontent.” The system discontent concerns democratic elements of politics such as parties, politicians, institutions, and the functioning of democracy. The élite discontent attacks the incumbent government and its performance in terms of its day-to-day policy outputs but also other parties including those in opposition (Hosch-Dayican, 2011).¹

To what extent was the M5S’s electoral success due to a general resentment against the established parties, the so-called casta (the clique)? Did it mainly refer to more specific and reasoned evaluations about political actors, their responsiveness, and their ability to handle crucial issues? Alternatively, did the voters cast their vote for M5S mainly because they were active, genuine supporters of the party? This question partially recalls the distinction between “instrumental” and “expressive” voting (Brennan and Hamlin, 1998). An instrumental voter is mainly led by the motivation of having an influence. The voter elaborates a rational, deliberate calculation of cost-benefits and, for this, needs some knowledge about party programs and policy positions. An expressive voter, on the other hand, would not necessarily be interested in the outcome of the election or be driven by specific parties’ policy positions, because benefits and costs derive from expressing support or opposition to the candidates/system as an end in itself (Brennan, 2001: 225). In that sense, the concept of expressive action incorporates also the dimension of belonging or identity. This implies that the focus has to be placed on political alienation rather than on policy preferences or attitudes, because protest voters are mobilized mainly by the need to express their grievances and are less interested in the program or policy positions of the party (Van der Brug et al., 2000; Van der Brug and Fennema, 2007). As a mean of expressing a lack of support for the government, the protest vote should concern all those parties perceived as being far from government influence, regardless of their ideological position (Bergh, 2004). In contrast, if a protest vote is read as instrumental, it stands for more than a generic punishment of established parties. Under this assumption, new parties would also gain “protest votes” by emphasizing some issues that are important to a growing and relevant share of voters (Van der Eijk and Franklin, 2009).

This means that voters may combine political dissatisfaction with ideological and issue considerations (Van der Brug, 1998), as in the case of extreme right or populist right parties which emphasize their negative attitudes toward immigration (Ignazi, 2003; Mudde, 2007; Rooduijn et al., 2014). In this case, the causal effect may be spurious, because the policy preferences are the real cause of an apparent protest effect and protest voters distrust the politicians only because voters see the issue differently (Bergh, 2004). A different position suggests that protest and issue evaluation might not necessarily be two sides of a coin. Ideological and pragmatic considerations improve (as opposed to weaken) an electoral preference driven by political dissatisfaction. For instance, the vote for a populist party can be read as more than a punishment of established parties but as rational instrumental behavior, reflecting clear reasoned consideration of specific political and social policies and how parties manage them (Hosch-Dayican, 2011). This argument leads to a different explanation of electoral behavior, provided by the theory of issue voting. In particular, the spatial model of voting (Rabinowitz and MacDonald, 1989) assumes that individuals compare the candidates or parties’ respective principles and vote for those whose political stances are closest to their own.

Data and methods

This article integrates two possible (and different) explanations for voting in favor of M5S in the 2013 election: one based on political discontent and the other based on policy preferences. Using bivariate and multivariate analysis (logistic regression models), our study aims to understand:

To what extent political discontent (and which kind of political discontent) explains the vote for M5S compared to other parties.

Whether the evaluations of those issues dominant in society and emphasized in the electoral campaign have counted in the individual decision to support the M5S.

How these evaluations interact with the dimension of discontent. In other words, whether including the dimension of issue evaluation—and in particular the voter–party proximity on certain issues—improve the explanatory power of a model based on the single impact of political protest, and which issue interacts better with protest.

We assume that (Hp1) the vote for the M5S in the 2013 general elections can largely be explained, more than for other parties, by the effects of political protest in both of its two components of (more radical) system distrust and (less radical) élites distrust. The M5S has strongly emphasized voters’ protest against the old parties, particularly the two largest ones, the Democratic Party (Partito Democratico (PD)) and The People of Freedom (Popolo della Libertà (PDL)), which in 2011–2013 were governing as a grand coalition.² In this way, the M5S had the opportunity to stress its willingness to present itself as a party that went beyond the left–right cleavage.

We also hypothesize that (Hp2) political protest has not acted as the only dimension influencing the electoral choice for M5S. Several studies (Biorcio and Natale 2013; Bordignon and Ceccarini, 2013; Corbetta and Gualmini, 2013; Itanes 2013; Passarelli et al., 2013; Tronconi, 2015) have showed that Grillo’s supporters and voters exhibit a peculiar profile: younger, highly educated, informed on social themes, savvy with technologies, less prone to be mobilized through traditional shortcuts such as party or left–right identification. Such elements suggest that protest should have counted as well as more articulated opinions on society and social issues. There has been a kind of additional effect led by the protest and reinforced by the issues dimension.

The data comes from a post-electoral survey (based on about 1500 face-to-face interviews) carried out 1 month after the Italian general elections of 2013 as part of a more general program of investigation into the electoral effects of the economic and political crisis in contemporary Italy, conducted by Itanes.³

The impact of protest in the vote for the Five Star Movement

A vote for a new (self-defined as antisystem) party such as M5S is expected to incorporate a certain amount of protest as an expression of disaffection with politics and/or, more specifically, with political actors and institutions. In order to correctly assess this influence, we adopt a classification of political protest that distinguishes two distinct dimensions of “system discontent” and “élite discontent” (Bergh, 2004; Hosch-Dayican, 2011). The protest against the political system is operationalized through a combination of variables related to certain feelings, which assume political relevance: (a) political trust/distrust, (b) efficacy/alienation, and (c) belief in a functioning party democracy.⁴ On the other hand, the protest against the political élite concerns: (a) perception of the state of the (personal/national) economy, (b) judgments on government performance and the attribution of responsibility for the economic crisis, and (c) opinion on leaders.⁵ In general, this kind of protest assumes, as a specific and focalized object of criticism, the incumbent party or parties, perceived as incapable of solving collective problems.

Tables 1 and 2 display, for the electorate as a whole, the general prevalence of a critical position along almost all the indicators, both those referring to the political institutions and those concerning how political actors worked. At the same time, the data show that political discontent is significantly higher among M5S voters than among supporters of any other party. This wider diffusion of negative feelings concerns, for instance, the dimension of “trust”: more than 40% of M5S voters display minimal (or zero) confidence in parliament and 60% of them have no trust in political parties, while the corresponding percentages for other voters are respectively around 20% and 30%. Moreover, M5S voters typically show little belief in the rule of parties as an instrument for participation or advocacy in democracy. In addition to this systemic discontent, Grillo’s supporters are also characterized by a very intense expression of discontent toward the political élites. In this scenario of generalized negative views, M5S voters express, therefore, the most critical evaluations about the economy (in particular, in terms of its perspectives) and about how the two governments of the XVI legislature (Silvio Berlusconi and Mario Monti) have faced up to the effects of the economic crisis. However, the most important peculiarity is probably that those attributions of negative responsibility among M5S voters turn out to be generalized to the entire political class, including the center-left opposition leader, whose rating is unexpectedly worse than the corresponding ratings attributed to the incumbent leaders. Furthermore, discontent among M5S supporters is not only higher when compared with other voters but also when the group of nonvoters is taken into account. This is a very important and innovative finding, because abstainers normally express distrustful, alienated, and critical positions.

Table 1.

Indicators of system discontent among M5S voters and other selected groups of voters.

	M5S voters	Centre-left (PD) voters	Centre-right (PDL) voters	Non voters	Cramers’ V (M5S vs. others)
Efficacy (% strongly agree)
People like me have no say in what the government does	56.6	45.5	51.8	61.8	0.078
Sometimes politics seems so complicated that a person like me cannot really understand what is going on	58.8	45.4	54.9	59.9	0.099
Those we elect lose touch with the people	72.6	57.2	62.5	64.0	0.123
Parties are only interested in people’s votes, not in their opinions	65.5	49.8	58.5	70.0	0.106
Trust in political institutions (% none)
Parliament	40.9	17.0	23.2	37.3	0.190
Parties	59.2	31.4	35.0	55.3	0.231
President of the Republic	18.5	2.8	6.7	12.4	0.256
European Union	20.5	4.9	11.5	20.7	0.218
Belief in functioning party democracy (% totally disagree)
Parties are necessary to protect the interests of different groups and social classes	24.2	10.0	14.6	15.5	0.195
Parties allow citizens to participate in political life	34.1	15.3	19.7	24.1	0.202
Without parties democracy would not exist	24.8	5.1	8.9	16.0	0.258
N	232	325	165	269

Source: authors’ elaboration from Itanes.

PD: Partito Democratico; PDL: Popolo della Libertà; M5S: Five Star Movement.

Table 2.

Indicators of élite discontent among M5S voters and other selected groups of voters.

	M5S voters	Centre-left (PD) voters	Centre-right (PDL) voters	Nonvoters	Cramers’ V (M5S vs. others)
Economic evaluations
State of the national economy, retrospective: very much worsened	68.5	57.8	64.5	61.2	0.109
State of the national economy, prospects: very much worsening	43.1	23.0	30.7	35.6	0.199
Relative deprivation: fear of losing job: very strong	14.7	6.6	4.9	7.0	0.158
Evaluation of government performance: % very negative
Berlusconi government on the economy	48.0	52.8	3.6	24.5	0.202
Monti government on the economy	28.9	12.0	36.4	28.2	0.184
Responsibility for economic crisis: % very responsible
Berlusconi government	24.2	23.6	0.0	10.8	0.169
Monti government	8.3	2.5	10.5	9.3	0.153
Leaders’ ratings: score 0–10, means
Centre-right colition leader: Berlusconi	1.5	1.1	7.4	3.3	0.243
Incumbent prime minister before 2013 elections: Monti	2.5	4.0	2.5	2.7	0.234
Centre-left coalition leader: Bersani	3.2	5.7	2.4	3.6	0.264
N	232	325	165	269

Source: authors’ elaboration from Itanes.

PD: Partito Democratico; PDL: Popolo della Libertà; M5S: Five Star Movement.

Political resentment assumes different meanings if expressed by highly engaged citizens or by citizens who feel that politics has little to do with their lives. In our analysis, we have taken into account this consideration by investigating the participative profile of different electorates. Findings in Table 3 show that M5S voters are more politically active and involved than center-right voters and nonvoters. The most interesting comparison is with Democratic party voters. In this case, M5S voters participate less than PD voters in conventional political participation but more in political mobilization, unconventional, or individual participation (send a letter, take part in a rally, boycott), while PD voters participate more in social activities. In terms of political attitudes, M5S voters are less interested and have less political knowledge, while their level of political information is higher in terms of online political information and lower in the case of traditional political information via TV news. In summary, those who have chosen M5S are, largely, engaged citizens who are responsive to a renewal repertoire of activism and activation. The large diffusion of political discontent among this electorate makes them similar to the “critical citizens” described by Norris (1999) and Dalton (1996).

Table 3.

Indicators of political involvement among M5S voters and other selected groups of voters.

	M5S voters	Centre-left (PD) voters	Centre-right (PDL) voters	Non voters	Cramers’ V (M5S vs. others)
Political participation	47.4	53.7	29.5	20.0	0.011
Party activism	4.3	8.9	4.2	1.5	0.048
Sign a petition	29.3	32.7	10.9	11.2	0.168
Send a letter	13.4	9.2	6.7	3.3	0.078
Attend public debate	25.0	26.9	17.5	6.7	0.040
Take part in rallies	17.7	12.9	5.4	5.2	0.076
Buy/boycott	17.7	14.1	4.2	5.6	0.081
Social participation	36.6	41.4	32.9	23.0	0.095
Interest in politics	41.6	50.8	27.2	17.7	0.064
Political knowledge	44.8	56.8	36.1	21.9	0.046
Offline information (reading)	55.7	52.2	38.0	27.7	0.107
Offline information (watching TV)	67.3	81.5	77.6	64.6	0.121
Online political information (at least one action)	62.9	49.7	40.0	37.5	0.188
N	232	325	165	269

Source: authors’ elaboration from Itanes.

Note: Reading article on politics during campaign: often+sometimes; Watching TV news: every day; Online political information: at least one action among those items: 1) surfing on party or candidate’s website, 2) surfing on party or candidate’s profile on social networks, 3) watching Tv campaign on Internet, 4) sharing campaign contents, 5) participate in online political discussions, 6) being invited online and participate in a political event.

PD: Partito Democratico; PDL: Popolo della Libertà; M5S: Five Star Movement.

Upon describing how protest is spread across the whole Italian electorate and particularly concentrated among M5S voters, we examine (Table 4) the impact of political discontent using a logistic regression, whose models test the (differential) effects, on the probability of voting for M5S, of the two additive indexes of “system discontent” and “élite discontent.” Model 1 (baseline) reports the coefficients for the main sociodemographic variables and the left–right self-placement dimension. With respect to this first step, a relevant finding is the considerable effect of age, with the youngest people likely oriented toward a vote for M5S. As expected, the left–right division does not discriminate, confirming the (self) representation of the party as a post-ideological actor able to drain votes from both of the ideological sides. The impact of the two components of political discontent has been tested in models 2 and 3. Confirming the results of bivariate analysis, regressions clearly show that all the expressions of resentment are relevant and significant, even after controlling the sociodemographic and ideological variables. Moreover, the maximum effect is reached when both the dimensions of discontent are included, as model 4 indicates.

Table 4.

Effect of political discontent on the probability of voting for M5S (logistic regressions).

	Model 1: baseline	Model 2: system discontent	Model 3: élite discontent	Model 4: system and elite discontent
Sex (female)	1.134	1.155	1.085	1.111
Age (<35 years)	2.408***	2.221***	2.450***	2.154***
Zone (south)	1.621**	1.121	1.508***	1.207
Education (low)	0.888	0.722	0.895	0.795
Occupational status (employed)	0.850	0.834	0.839	−0.138
Left–right position (scale, 0–1)		1.070	1.082	1.077
System discontent (min–max, scale 0–1)		2.692
Élite discontent (min–max, 0–1)			3.177***
System discontent + elite discontent (min–max, scale 0–1)				4.190***
Constant	0.215***	0.039***	0.244***	0.019***
Pseudo R ²	0.033	0.076	0.087	0.116
LR χ ² (7), (p)	37.93 (0.000)	83.24 (0.000)	94.54 (0.000)	116.49 (0.000)
Hosmer and Lemeshow, χ ² (8), (p)	12.44 (0.132)	9.86 (0.275)	5.59 (0.693)	5.19 (0.736)
N	927	824	927	815

Source: authors’ elaboration from Itanes.

Note: the dependent variable is coded 0 (vote for other parties) and 1 (vote for M5S). The two components of political protest – system discontent and élite discontent – are additive indexes consisting of all the variables in Tab. 1 (system discontent) and Tab. 2 (élite discontent). Both the indexes are normalised. Coefficient are reported as odds ratios *p < 0.01; **p < 0.005; ***p < 0.001.

M5S: Five Star Movement.

The importance of the negative feelings toward politics/politicians as an explanation of voters’ choices in the 2013 parliamentary elections is clearly visible in Figure 1, which reports the predicted percentages of votes for the main parties under the effect of lower/higher political discontent (an additive index of system discontent and élite discontent). While the predicted values for PD and PDL significantly decline according to the growth of political discontent, the trend for M5S is inverse. The probability of attracting voters who are not critical or only moderately critical toward politics is lower than those expressed by other parties. In the absence of criticism, voters mainly choose the PD, to some extent the PDL, and only minimally M5S. However, when the level of criticism among the electorate grows, the predicted percentages of votes for M5S are very high and the M5S becomes the most preferred party.

Figure 1.

Predicted percentage of voting for M5S, PD, and PDL by intensity of political discontent. PD: Partito Democratico; PDL: Popolo della Libertà; M5S: Five Star Movement.

A reasoned vote supporting the protest? The influence of issue evaluations

In the previous paragraph, we have pointed out that the dimension of “political discontent” in all its components acted as a crucial predictor of voting for M5S. We proceed further along this line of investigation, seeking to clarify under which terms and to what extent voting for M5S has been led by a pure effect of refusal and poor evaluation of the political class and institutions, or whether it depicts a more complex articulation encompassing other factors and explanations, acting in combination with negative feelings toward politics (Conti and Memoli, 2015).

In order to answer this question, we have introduced to the analysis a new and different set of indicators concerning policy preferences. Our data set covers four domains of policy: EU integration (less vs. more EU integration), private versus public (less taxation vs. more public services), families (less vs. more rights for nontraditional families: contrary vs. in favor), and immigration (immigrants are too much: in favor vs. contrary on this statement). The left–right self-placement is also included as a super issue incorporating more general information on the ideological position. We have taken into account two different kinds of information: the individual position on a single issue and the proximity/distance between the voter and the (perceived) party’s position on that issue. The former dimension simply tells us where M5S voters place themselves on a specific relevant social problem debated in the electoral campaign and whether their positions significantly differ from other voters’ positions. More relevant for our analysis are the measures of voter–party distance. As the spatial model of voting suggests, a short distance (in other words, high or complete proximity) means that voters are aligned with the party on a political stance and this closeness may have acted as an incentive in favor of a vote for that party.

Table 5 gives a general overview of the profiles of different electorates on the policy preferences described above. M5S voters hold the most extreme position in only one of these issues, “EU integration” (more contrary), while their position is close to (as favorable as) that of PD voters on the issue “the right for new families.” On the remaining issues, “less tax-more public services” and “immigration,” M5S voters place themselves between the positions of other electorates of the center-left and center-right. Regarding the other information on voter–party distance, the average score reaches about 1–2 points difference (along this scale with minimum 0 and maximum 6 points distance). Political distance on left–right is more pronounced among M5S voters (1.47; PD: 1.01, PDL: 0.74), but the heterogeneity is not confirmed when other issues are taken into account. In general, the level of distance is in line with those expressed by PD and PDL electorates. The scores are higher on “EU integration” and “immigration,” issues that divide other voters too. M5S voters seem to be more aligned with their party on the issue “right for new families.” Moreover, in any case, party–voter distance is the highest.

Table 5.

Position on issues and issue distance for M5S, PD, and PDL voters (means).

	M5S voters	Centre-left (PD) voters	Centre-right (PDL) voters
Voters position on issues, means
EU integration (cons vs. pro, scale 0–1)	0.38	0.55	0.41
More public services (cons vs. pro, scale 0–1)	0.43	0.50	0.34
Rights for new families (cons vs. pro, scale 0–1)	0.54	0.55	0.30
More immigrants needed (cons vs. pro, scale 0–1)	0.25	0.33	0.12
Ideological position: left–right self-placement (0–1, max right)	0.40	0.24	0.78
Voters–party absolute distance
EU integration	1.40	1.34	1.51
More public services	1.19	1.21	1.14
Rights for new families	1.00	1.29	1.20
More immigrants needed	1.56	1.74	1.08
Ideological position, left–right	1.47	1.01	0.74

Source: authors’ elaboration from Itanes.

Note: distance is calculated as absolute difference between voter’s position and perceived party’s position. Each policy preference is registered as position on a 1-7 scale. Issue distance covers a range of maximum 6 points (0-6).

PD: Partito Democratico; PDL: Popolo della Libertà; M5S: Five Star Movement; EU: European Union.

Regardless of the debate about the post-ideological and protest-centered vote, empirical evidence (Table 5) suggest that M5S voters’ positions on the traditional issues discussed in the electoral campaign are not so different from what other voters indicate. At the same time, they do not perceive a clear discrepancy or distance from their (self-perceived) party’s position on the same argument. We now move to a final step, namely attempting to redefine the explanatory model of voting for M5S. A crucial question in our analysis concerns the relative strength of different explanations of voting for M5S: Do voters choose M5S because of their generic (and polemic) evaluation on politics and elites or do they also attribute importance to how (much) the party is perceived as closer to their position on a given topic? To assess the importance of policy preferences and of how they combine with criticism, we run a regression analysis where models include political discontent and the interaction of this variable with voter–party indexes of distance/proximity on different issues. If voters mostly agree with the party on policy(ies) consideration, voters’ choice reflects idealistic voting (Bergh, 2004; Van Der Brug et al., 2000) based on arguments other than the simple expression of discontent and related to a concordance of positions on one or more topics viewed as important or salient. There are three possible scenarios in which issue voting and protest voting combine. As first, interaction may assume a positive sign: the closer the issue position of the party to the voter’s position, the stronger will be the effect of protest on voting for that party. Dissatisfaction with the system and/or elite performance might have a stronger effect when it is coupled with ideological or issue preferences. A low distance (proximity) combines positively with protest and the magnitude of the effect is higher, in a sort of “reasoned protest voting” where critical voters decide to choose M5S not only because of their criticism but also because they agree with most of this party on the issue/s they consider as important. On the other hand, we could expect that policy preferences have acted instead of protest, giving an alternative motivation to those voters less (or not at all) driven by political discontent. In this case, interaction will be negative: the probability of voting for the party rises in the absence (not in the presence) of political resentment. As a further possible scenario, it could be the case that high voter–party proximity on issues does not change in any way the intensity and direction of the existing effect of political discontent.

The analysis reported in Table 6 mostly supports the assumption of a positive interactive effect. We run five logistic regression models. Model 1 only includes political discontent without any interaction. As described in Table 4, the effect of this composite variable is strongly positive: the more negative the political feelings, the greater the probability of voting for M5S. Models 2–5 add interactive effects between political discontent and the measures of voter–party issue distance/proximity on each policy preference. We have recoded issue distance/proximity as a 0–1 dichotomy, where 1 corresponds to null distance between voter and party; in other words, the maximum level of proximity. Findings give us precise indications on the impact of all distinct interaction included in the models. Coefficients (odds ratios) are higher than 1 and assume significant values when discontent combines with at least two of the four issues: proximity on “EU integration” and proximity on “Rights for new families”. In fact, models 2 and 4 improve the goodness of fit, as displayed in Likelihood Ratio (LR) χ ² statistics.

Table 6.

Effect of political discontent and issue proximity on voting for M5S (logistic regression).

	Model 1	Model 2	Model 3	Model 4	Model 5
Sex	1.111	1.129	1.128	1.119	1.115
Age (<35 years)	2.154***	2.203***	2.154***	2.120***	2.151***
Education	0.795	0.801	0.780	0.814	0.796
Zone	1.207	1.206	1.245	1.201	1.237
Occupational status	0.870	0.879	0.855	0.849	0.843
Left–right position	1.077	1.076	1.077	1.083	1.071
Political discontent	4.190***	3.617***	3.630***	2.499***	3.284***
EU integration		0.776
EU integration × Discontent		1.881*
Private vs. public			0.759
Private–public × Discontent			1.349
New families				0.342
New families × Discontent				2.666***
Need of immigration					0.397
Need of immigration × Discontent					1.409
Constant	0.019***	0.206***	0.021***	0.029***	0.028***
Pseudo R ²	0.116	0.119	0.117	0.132	0.118
LR χ ² (p)		3.45 (0.178)	1.90 (0.388)	14.14 (0.000)	3.00 (0.223)
Hosmer and Lemeshow, χ ² (p)		5.84 (0.664)	5.86 (0.663)	11.43 (0.178)	11.82 (0.159)
N	815	815	815	815	815

Source: authors’ elaboration from Itanes.

Note: Political discontent index is a dichotomy and assume value 0-1 (high discontent). Issue variables (EU integration, Private vs. Public, New families, Need of immigrant) are dichotomized into 0-1 with 1 = proximity (no distance between voter and party’s position). LR (likelihood ratio) chi-square test compare the goodness of fit of Models 2-5 compared with Model 1. Coefficient are reported as odds ratios. *p < 0.01; **p < 0.005; ***p < 0.001.

M5S: Five Star Movement; EU: European Union.

Finally, we have reproduced these results (Figure 2) as predicted probabilities of voting for M5S according to the level of political discontent. All curves drawn in the figure correspond to the regression models reported in Table 6: one with political discontent alone, the other four including the effect of political discontent plus a single issue. The level of predicted voting for M5S is slightly higher, compared with model 1, when political discontent combines with voter–party closeness on a policy. This interaction effect is found to be more pronounced when voters share the party’s position on the issues on EU integration and the rights for new families and less pronounced in the other two cases. Hypothesis 2 is partially confirmed. The explanation of voting based on issue evaluations does not count as well as the explanation based on political discontent, but these two models are not alternative. Having a shared position with the party on a specific policy preference (in particular, on two policies such as EU integration and the right for new families) does not boost the probability of voting for M5S when discontent is low. Instead, a multiplicative effect significantly occurs when voters show a medium and, with a stronger intensity, higher level of discontent, indicating that there a sort of reinforcing effect is still in play, very different from a situation where issue counts as an alternative to protest and protest-oriented voters do not use information on issues but only their distrust to decide.

Figure 2.

Predicted probability of voting for M5S by level of political discontent; only political discontent and political discontent and issue proximity. M5S: Five Star Movement.

Conclusion

Protest voting has been often overlapped with the concept of populism. The two lemmas refer to different political phenomena, which only sometimes are evident simultaneously, and there is no unique and agreed definition of such a phenomenon. In literature, the “protest vote” is generally considered to be a vote primarily cast to scare the élite but also a vote for which political attitudes are expected to be of minor importance. The idea of protest voting necessitates something against which voters are protesting, some kind of disapproval and is the result of political distrust. A vote for nonestablished parties incorporates, in its antielitist position, a certain amount of dissatisfaction, and it is expected to frighten or shock the élites from mainstream parties. Protest voting may be defined as the level of dissatisfaction with how the established parties and the system, as well as the government and the parliament, have performed.

New parties, not only in Italy, have perhaps been too quickly labeled as protest parties. In order to go beyond impressionist views, we have investigated to what extent the electoral success for the M5S in 2013 Italian elections was an expression of political dissatisfaction and whether more general evaluations on social issues have counted in interacting with dissatisfaction. This was relevant given that the promotion of voters’ political awareness and attentiveness was an element explicitly enhanced during the M5S’ electoral campaign. Findings partially support this latter explanation. The probability of voting for Grillo’s party significantly increases when high levels of political dissatisfaction combine with voter–party proximity on issues, in particular when policy preferences concerning EU integration and the rights for new families are taken into account. Political protest matters in voting behavior, but the groundbreaking result in the case of the M5S is that dissatisfaction with the system and/or elite performance has a stronger effect when coupled with ideological or issue preferences. Our findings confirm the hypothesis of a reinforcement of two different effects, and they add further considerations on the dimension that has counted more. It is not merely a proximity on a typical issue emphasized by populist parties such as immigration but the presence of a shared position on an unexpected issue, generally supported by parties other than populist ones. M5S’ voters are more prone to claim right for new families as well as contrast the European integration, a manifest sign of the across-the-board electorate. Although those issues incorporate different ideological positions, they well combine with the protest dimension and are predictors of mobilization in favor of M5S. In general, policy preferences seem to have not counted for those voters without an intense protest attitude, while they improve the explicative effect of protest. We cannot assume that voting for the M5S does not exist beyond and without political dissatisfaction. The point is that the protest effect is amplified by the issues, which enlarge the spectrum and the probability of success. The M5S is a party that groups together a strong protest, which is not generic but reasoned, oriented by a political awareness on important issues. Therefore, our study fits with a number of other research projects, in particular those that have discussed and integrated protest and issue voting.

Footnotes

Acknowledgements

We are grateful for the comments we received from the two anonymous reviewers that contributed to improve our article. We also thank for the comments on earlier drafts from PaoloBellucci, Piero Ignazi, Marco Maraffi, Paolo Segatti, and participants in seminar at University of Siena in 2015 on “Leaders without a base? Parties (and voters) in Italy” promoted by the Standing Group on Political Parties Opinion and Voters of the Italian Political Science Association.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Albertazzi

McDonnell

(2015) Populist in Power. London: Routledge.

Bergh

(2004) Protest voting in Austria, Denmark and Norway. Scandinavian Political Studies 27(4): 367–389.

Almond

Powell

(1966) Comparative Politics. A Developmental Approach. Boston: Brown and Co.

Barnes

Kaase

Allerback

. (1979) Political Action: Mass Participation in Five Western Democracies. Beverly–Hills-London: Sage.

Biorcio

Natale

(2013) Politica a 5 stelle. Milano: Feltrinelli.

Bordignon

Ceccarini

(2013) Five stars and a cricket. Beppe Grillo. Shakes Italian politics. South European Society and Politics 18(4): 427–449.

Bowler

Lanoue

(1992) Strategic and protest voting for third parties: the case of the Canadian NDP. Political Research Quarterly 45(2): 485–499.

Brennan

(2001) Five rational actor accounts of the welfare state. Kyklos 54(2–3): 213–234.

Brennan

Hamlin

(1998) Expressive voting and electoral equilibrium. Public Choice 95(1–2): 149–175.

10.

Canovan

(1981) Populism. New York: Harcourt Brace. Jovanovich.

11.

Conti

Memoli

(2015) The emergence of a new party in the Italian party system: rise and fortunes of the Five Star Movement. West European Politics 39(3): 516–534.

12.

Corbetta

Gualmini

(2013) Il Partito di Grillo. Bologna: Il Mulino.

13.

Dalton

(1996) Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies. New York: Chatham House.

14.

De Luca

(1995) The Two Faces of Political Apathy. Philadelphia: Temple University Press.

15.

Hermet

(2001) Le Populismes Dans le Monde. Une Histoire Sociologique. XIX^e-XX^e Siècle. Paris: Fayard.

16.

Hosch-Dayican

(2011) Populist Party Choice in Germany and the Netherlands. The Protest Vote Hypothesis Revisited. 10th Dutch-Belgian Political Science Conference, Amsterdam, 9–10 June 2011.

17.

Ignazi

(1992) The silent counter-revolution. Hypotheses on the emergence of extreme right parties in Europe. European Journal of Political Research 22(1): 3–34.

18.

Ignazi

(1996) The crisis of parties and the rise of new political parties. Party Politics 2(4): 549–566.

19.

Ignazi

(2003) Extreme Right Parties in Western Europe. Oxford: Oxford University Press.

20.

Itanes (2008) Il Ritorno di Berlusconi. Bologna: Il Mulino.

21.

Itanes (2013) Voto Amaro: Disincanto e Crisi Economica Nelle Elezioni Del 2013. Bologna: Il Mulino.

22.

Kang

(2004) Protest voting and abstention under plurality rule elections. An alternative public choice approach. Journal of Theoretical Politics 16(1): 79–102.

23.

Kselman

Niou

(2011) Protest voting in plurality elections: a theory of voter signalling. Public Choice 148(3–4): 395–418.

24.

Kriesi

Pappas

(eds) (2015) European Populism in the Shadows of the Great Recession. Colchester: ECPR Press.

25.

Kriesi

(2012) The political consequences of the financial and economic crisis in Europe: electoral punishment and popular protest. Swiss Political Science Review 18(4): 518–522.

26.

Mudde

(2007) Populist Radical Right Parties in Europe. Cambridge: Cambridge University Press.

27.

Norris

(1999) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press.

28.

Opp

(2009) Theories of Political Protest and Social Movements: A Multidisciplinary Introduction, Critique, and Synthesis. Abingdon: Routledge.

29.

Passarelli

(2016) Populists and the Lega Nord. In: Jones

Pasquino

(eds) The Oxford Handbook of Italian Politics. Oxford: Oxford University Press, pp. 224–239.

30.

Passarelli

Tronconi

Tuorto

(2013) Dentro il movimento. Organizzazione, attivisti e programmi. In: Corbetta

Gualmini

(eds) Il Partito di Grillo. Bologna: Il Mulino, pp. 123–167.

31.

Passarelli

Tuorto

(2012) Lega & Padania. Storie e Luoghi Delle Camicie Verdi. Bologna: Il Mulino.

32.

Passarelli

Tuorto

(2013) Berlusconi tra elettori fedeli e defezioni. In: ITANES (eds) Voto Amaro. Disincanto e Crisi Economica Nelle Elezioni Del 2013. Bologna: Il Mulino, pp. 71–82.

33.

Passarelli

Tuorto

(2014) Not with my vote: turnout and economic crisis in Italy. Contemporary Italian Politics 6(2): 147–158.

34.

Pedersen

(1982) Towards a new typology of party lifespans and minor parties. Scandinavian Political Studies 5(1): 1–16.

35.

Poguntke

(1987) New politics and party systems: the emergence of a new type of party? West European Politics 10(1): 76–88.

36.

Rabinowitz

MacDonald

(1989) A directional theory of issue voting. American Political Science Review 83(1): 93–121.

37.

Raniolo

(2006) Forza Italia: a leader with a party. South European Society and Politics 11(3–4): 439–445.

38.

Rooduijn

de Lange

van der Brug

(2014) A populist Zeitgeist? Programmatic contagion by populist parties in Western Europe. Party Politics 20(4): 563–575.

39.

Sartori

(1976) Parties and Party Systems. Cambridge: Cambridge University Press.

40.

Tarchi

(2003) L’Italia populista. Dal Qualunquismo ai Girotondi. Bologna: Il Mulino.

41.

Tronconi

(2015) Beppe Grillo’s Five Star Movement. Farnham: Ashgate.

42.

Van Biezen

(2003) Political Parties in New Democracies: Party Organization in Southern and East-Central Europe. London and New York: Palgrave Macmillan.

43.

Van der Brug

(1998) The informed electorate: political perceptions and party behaviour. Acta Politica 33(1): 20–55.

44.

Van der Brug

Fennema

(2007) What causes people to vote for a radical-right party? a review of recent work. International Journal of Public Opinion Research 19(4): 474–487.

45.

Van der Brug

Fennema

Tillie

(2000) Anti-immigrant parties in europe: ideological or protest vote? European Journal of Political Research 37(1): 77–102.

46.

Van der Eijk

Franklin

(eds) (1996) Choosing Europe? The European Electorate and National Politics in the Face of Union. Ann Arbor: University of Michigan Press.

47.

Van der Eijk

Franklin

(2009) Elections and Voters. Basingstoke: Palgrave Macmillan.

48.

Van Kessel

(2015) Populist Parties in Europe: Agents of Discontent? New York: Palgrave Macmillan.

49.

Vegetti

Poletti

Segatti

(2013) When responsibility is blurred: Italian national elections in times of economic crisis, technocratic government, and ever-growing populism. Italian Political Science Review 43(3): 329–352.