ArticlePDF Available

The neo-fascist territorial legacy and the success of the Lega in the 2019 European elections: a multilevel approach


Abstract and Figures

In Italy, the Lega obtained outstanding electoral success in the 2019 European elections, becoming the first party on the political spectrum. Previous literature has argued that this performance can be attributed to the leadership of Matteo Salvini, who transformed the Lega from an ethno-regionalist party into a national right-wing party (Passarelli and Tuorto, 2018). Previous research has also argued that the recent geographical trajectories of the party's success might be associated with the prevalence of a neo-fascist minority during the First Republic (e.g. Mancosu, 2015). However, the empirical evidence comes from aggregate official results and focuses only on some specific Italian regions of the so-called 'red-zone'. By employing multilevel models on survey data, this paper tests whether this expectation holds also at the individual level, and in a larger geographical area. The findings show that individual propensities to vote for the Lega in 2019 are associated with the percentage of votes obtained more than forty years ago by the Movimento Sociale Italiano in the municipality where the respondent lives, but only in central and southern Italian regions, in which the Lega was an irrelevant competitor before Salvini's leadership. These findings provide additional evidence concerning the ideological drivers of preferences for the Lega.
Content may be subject to copyright.
Italian Political Science, VOLUME 14 ISSUE 2, OCTOBER 2019
© 2019 Italian Political Science. ISSN 2420-8434.
Volume 14, Issue 2, x-x.
Contact Author: Moreno Mancosu, University of Turin.
E-mail address:
The neo-fascist territorial legacy and
the success of the Lega in the 2019
European elections: a multilevel approach
Moreno Mancosu
Riccardo Ladini
In Italy, the Lega obtained outstanding electoral success in the 2019 European elections, becoming the first
party on the political spectrum. Previous literature has argued that this performance can be attributed to the
leadership of Matteo Salvini, who transformed the Lega from an ethno-regionalist party into a national right-wing
party (Passarelli and Tuorto, 2018). Previous research has also argued that the recent geographical trajectories
of the party’s success might be associated with the prevalence of a neo-fascist minority during the First Republic
(e.g. Mancosu, 2015). However, the empirical evidence comes from aggregate official results and focuses only
on some specific Italian regions of the so-called ‘red-zone’. By employing multilevel models on survey data, this
paper tests whether this expectation holds also at the individual level, and in a larger geographical area. The
findings show that individual propensities to vote for the Lega in 2019 are associated with the percentage of
votes obtained more than forty years ago by the Movimento Sociale Italiano in the municipality where the re-
spondent lives, but only in central and southern Italian regions, in which the Lega was an irrelevant competitor
before Salvini’s leadership. These findings provide additional evidence concerning the ideological drivers of pref-
erences for the Lega.
1. Introduction
n Italy, the most relevant outcome of the 2019 European Elections was the outstand-
ing consensus for the Lega, which obtained more than 9 million votes, equal to 34% of
valid votes. Since its foundation about thirty years ago, the Lega, for the first time, has
become the most supported party on the political landscape at the national level. The re-
sult is even more stunning when considering the partys strong ethno-regionalist
background, rooted in the northern part of the country. We should also consider that the
Legas best performance at the national level before the leadership of Matteo Salvini, who
became leader of the party in December 2013, was far lower than the most recent one
dating back to the 1996 national elections, when 3.8 million people voted for Lega Nord
(about 10% of valid votes).
Until a few years ago, the electoral performance of the Lega could be depicted by
the metaphor of an expanding and contracting lung, always confined to the northern
regions of the country (Segatti and Vezzoni, 2011). In political times characterized by
The neo-fascist territorial legacy and the success of the Lega
the relative stability of the political system, the Lega reduced its area of consensus and
maintained some support only around its strongholds. In more unstable political pe-
riods (e.g. the 1992 and 1996 elections, characterized by significant political turmoil),
support for the Lega overcame its traditional boundaries by reaching some peripheral
areas of its territorial basin, such as some provinces of Emilia-Romagna. In the last
electoral cycle, pundits and academics have observed the increasing nationalization
of the Lega, which largely increased its consensus in the red area in the 2015 regional
elections and, starting from the 2018 national elections, became a relevant political
actor even in central and southern regions.
This paper relies on the strand of literature that posits the relevance of mecha-
nisms of behavioural path dependence to understanding social and political facts (e.g.
Diamanti, 2003; Acharya et al., 2018). According to this literature, individual atti-
tudes and behaviours are partially related to a self-reinforcing set of geographically-
located behaviours and attitudes that persist over time. Similarly to institutional path
dependence (Putnam et al., 1993), behavioural path dependence literature argues that
attitudes are passed down through generations, by means of grass-roots learning
mechanisms (in this respect, the role of primary socialization is crucial) and by local
social structures (schools, churches, associations, etc.) that persist in the territory
(Sani, 1976; Wittenberg, 2006; Voigtländer and Voth, 2012; Acharya et al., 2018). As
a result, we are able to trace a certain political culture defined geographically back
to its roots by identifying the characteristics of the same geographical place several
decades (or even centuries) before.
Scholars have stressed that the success of the Lega is, among other things, related to
the rapid change in the party agenda enacted by the new leader, Matteo Salvini, who
aimed to transform what was originally an ethno-regionalist party into an extreme right-
wing national one (Passarelli and Tuorto, 2018). In this regard, the Lega started adopting
a totally new set of claims and communication strategies, sometimes appealing to a nar-
rative that resonates with that of the neo-fascist parties of the First Republic (above all,
the Movimento Sociale Italiano, MSI). According to previous studies (Mancosu, 2015;
Mancosu and Ladini, 2018), in regions in which the Lega has never obtained a sufficient
level of support because of its former ethno-regionalist structure, this new more ex-
treme and nationalist rhetoric might have been particularly appealing for voters living
in areas with a stronger neo-fascist political tradition. The present paper aims to test the
patterns of geographical diffusion of the newLega in central and southern Italian re-
gions by providing, among others, a contextual-level explanation of the success of
Salvinis Lega.
We combine survey data at the individual level with official data at the contextual
level to show that, in the centre-south of Italy, individual propensities to vote for the Lega
during the 2019 election campaign are positively correlated with the electoral perfor-
mance of the Movimento Sociale Italiano (measured more than forty years ago) in the
municipality where the respondents live.
The paper is structured as follows: the next section focuses on the success of the Lega
in the 2019 European elections beyond its traditional borders, illustrates the shift in the
communication strategy of the party, and describes the possible social mechanisms fos-
tering the partys diffusion trajectories. The third section explains data, variables and
models employed in the paper; the fourth section presents the results and the fifth draws
some conclusions and suggests some limitations of the paper.
2. Background
2.1. The Lega beyond its borders
Previous studies have provided various explanations for the increased performance of the
Lega on the political spectrum and the overcoming of its traditional borders (Mancosu,
2015; Albertazzi et al., 2018; Mancosu and Ladini, 2018; Passarelli and Tuorto, 2018).
Nonetheless, these explanations agree that the success of the Lega is largely dependent
on the shift in the partys rhetoric carried out by Matteo Salvini. Since he took office, the
new leader of the party has characterized his political action by transforming the Lega
from a somewhat traditional right-wing regionalist party (Tarchi, 1998), into a na-
tional/nationalist right-wing party (Albertazzi et al., 2018; Passarelli and Tuorto, 2018)
radically based on anti-immigrant and anti-EU rhetoric.
As outlined by Passarelli and Tuorto (2018, 90), although the Lega started to move
towards the right in the mid-1990s, the leadership of Salvini sped upthe process of rad-
icalization of the party. This strategy was intended to increase the potential electoral
basin of the Lega, with the realization that, among the Italian population, anti-immigra-
tion attitudes are widespread (Sniderman et al., 2002; Vezzoni, 2018) and anti-EU
attitudes have substantially increased in the last 20 years (see Albertazzi, 2016; Brunazzo
and Della Sala, 2016; Serricchio, 2018).
Moreover, the refugee crisis which led to the arrival of about 600,000 immigrants
in Italy from 2014 to 2017 contributed to making the voice of the Lega louder on the im-
migration issue. In this respect, we should outline that the salience of this issue
dramatically increased between 2013, when only 4% of Italians considered it as the first
or second Italian problem, and 2018, when the percentage rose to 23% (Vezzoni, 2018).
By shifting its main enemy from Rome to Brussels (Brunazzo and Gilbert, 2017) and
avoiding attacks against southern Italians (largely present in the first incarnation of the
party, Lega Nord), Salvinis Lega has been successful in going beyond its traditional geo-
graphical boundaries.
Starting from the last election of the pre-Salvini era the 2013 national elections in
which the share of votes for the Lega was around 4% the positive trend of party support
has led the Lega to largely increase its consensus at every election (obtaining 6% of valid
votes in the 2014 European elections, 17% during the 2018 national elections, and 34% in
the 2019 European elections). To appreciate the effectiveness of the Legas strategy, Ta-
ble 1 compares the percentages of valid votes for the Lega before and after the
appointment of Matteo Salvini as leader of the party in two European elections (2019 and
2009 in which, incidentally, party performance was above the historical average up to
that moment).
The five areas considered refer to the Istat geographical categorization, where the northwest is com-
posed of Valle dAosta, Liguria, Piemonte and Lombardia, northeast includes Veneto, Friuli, Trentino-
Alto Adige and Emilia-Romagna, Centre includes Toscana, Marche, Umbria and Lazio, South includes
Abruzzo, Molise, Puglia, Basilicata, Campania, and Calabria, and Islands comprise Sicily and Sardegna.
We dec ided not to employ the geopolitic al categorizati on (Galli, 1968), in order to keep separ ate Emilia
The neo-fascist territorial legacy and the success of the Lega
Table 1. Voting percentages for the Lega in 2009 and 2019 European Elections in Italy
Lega Nord 2009
Lega 2019
Geographical area
In addition to the partys growth in its strongholds (in which the Lega doubled its
strength in the northeast and northwest), the most stunning performance is the increase
in support in places in which the party was substantially irrelevant 10 years ago (becom-
ing 11 times larger in the centre, 39 times in the south and 56 times in the islands). We
can thus say that the 2019 European elections consolidated the new geographical pattern
which started at the beginning of Salvinis office.
2.2. A neo-fascist inheritance?
Why did this operation become so effective? The new era of the Lega comes in a moment
of crisis of the centre-right coalition, which began with the economic crisis of 2011. The
progressive erosion of the image and political credibility of the main shareholder of the
centre-right coalition, Silvio Berlusconi, and his personal party Forza Italia, combined
with the corruption scandal that involved the former leader of Lega Nord, Umberto Bossi,
in 2012, allowed new actors to exploit the potential of this political space.
If, on the one hand, the opening of this political space has been crucial in the new
success of the Lega, on the other it is also important to understand what kind of claims
and communication strategies filled this political gap. Salvinis communication, since
the very beginning of his office, has been aggressive and mainly based on negative cam-
paigning (Bracciale and Martella, 2017), aimed at targeting a certain number of enemies
of his potential electoral constituency (immigrants, NGOs, intellectuals, and everything
that might recall the leftor the outgroup) and repeatedly attacking them on every me-
dium at his disposal.
The communication strategy enacted by the new leadership was also characterized
by a certain hiddenopenness to even more extreme positions, a form of scantily dis-
guised sympathy towards more extreme right-wing parties and, in general, to the neo-
fascist area. For instance, as noticed by De Giorgi and Tronconi (2018, 341), While the
use of violence by [...] extreme right groups was apparently condemned, [...] the League
[...] often justified episodes of racism and xenophobia that took place during the course of
2017, placing the blame ultimately on the various governments in office over recent years
and the decisions they took on immigration. Other examples can be traced back by em-
ploying the direct communication of its leader via social media, often characterized by
Romagna, where the Lega was a relevant competitor even before Salvinis leadership (Passarelli and
Tuorto, 2012, 2018), and the other regions belonging to the red zone, where the Lega obtained a very low
consensus in the pre-Salvini period (see Mancosu and Ladini, 2018).
positive referencing to the public works done during the so-called Ventennio(the 1922-
1943 period of fascist dictatorship, see Rame, 2018). Although this communication style
could be interpreted as a form of tactical dog-whistle politics(Haney-Lòpez, 2014), in-
stead of a clear endorsement of the dictatorship, the subtle but insistent references to the
fascist regime and its most important characters might have produced some effect on the
resilient minority of Italian voters that look with sympathy and nostalgia on the Venten-
nio period as a golden ageof order, discipline and security (see Castelli Gattinara, 2018).
This shift toward the right in the rhetoric of the party was also reflected in the profile of
the voters of the Lega. According to the Itanes data, while in the mid-1990s, on average,
voters of Lega Nord located themselves at the centre of the left-right (0-10) scale, their
average values increased from 6.8 to 7.8 from 2001 to 2013, and to 8.2 in 2018 (Passarelli
and Tuorto, 2018).
In this respect, several studies have argued that the Lega diffusion trajectories in the
last 5 years might be partially explained by the more extremist and nationalist turn oper-
ated by Salvini. In some of these studies (Mancosu, 2015; Mancosu and Ladini, 2018), it
is shown, by means of aggregate analyses, that places in which the performance of the
newLega was stronger are also those contexts in which the neo-fascist minority, meas-
ured by means of the support for the Movimento Sociale Italiano (MSI) in the 70s, was
strong as well.
As concerns the social mechanism that might explain this empirical evidence, we
move in the wake of the strand of literature aimed at explaining political and electoral be-
haviour by means of patterns of behavioural path dependence (Acharya et al., 2018).
Analogously to path dependence in institutions (see, above all, Putnam et al., 1993) this
strand of literature argues that attitudes and behaviours become, in some way, self-rein-
forcing: once a path of development is embedded in the political culture of a certain
geographical space, it becomes increasingly harder to erase. Similarly to institutions like
language, religion and rituals, attitudes are passed down from generation to generation,
encouraged by families and mechanisms of primary socialization and by social struc-
tures, such as schools, churches, political and non-political institutions rooted in the
territory (see Acharya et al., 2018). Many examples which suggest behavioural path de-
pendence can be traced back in the literature: Voigtländer and Voth (2012) show that the
Nazi Partys support in 1933 elections was higher in places that had Jewish pogroms dur-
ing the fourteenth century (showing a persistence in behaviour of almost 500 years);
Wittenberg (2006), by analyzing voting behaviour in Hungary before and after the com-
munist period, shows that the performance of right-wing parties after communism was
higher in areas in which right-wing parties had higher consensus during the pre-com-
munist period.
The social mechanism argued in these studies is somewhat similar to theories ex-
plaining the persistence over time of scattered zones of political influence in Italy: areas
with a strong communist or Christian-democratic tradition during the First Republic
were usually characterized by a stronger activism of militants of the local majority party,
as well as a stronger presence of associations connected to the party itself (Bellucci and
Segatti, 2010). Moreover, in these areas, it was more likely that the primary socialization,
operated by the family of origin, was more able to convey a precise set of values compatible
with the ideology of the dominant party.
The neo-fascist territorial legacy and the success of the Lega
In our case, the mechanisms argued here have a probabilistic component. Given cer-
tain contextual conditions, a larger quota of the local electoral body will be more likely to
accept a neo-fascist rhetoric if the characteristics of the context are exogenously more
permeable to such rhetoric. What has been argued in the previous literature, thus, is that
the shift of the communication operated by Salvini could have contributed to triggering
parts of the electoral body that were exogenously prone to accepting this type of rhetoric
(and that one of the more prompt measures to identify the exogenous propensity to accept
it is the prevalence of the neo-fascist right in a previous period). We need to outline that
since support for the MSI during the First Republic was far lower than that for the
Democrazia Cristiana and Partito Comunista Italiano, its political tradition could be re-
ferred to as a minority political tradition(Mancosu and Ladini, 2018). Moving from the
suggestion of Sani (1976, 394), who argued that the search for contextual forces shaping
political attitudes should not be limited only to the identification of the major political
tradition of the larger community, our contribution aims at analysing how neo-fascist
political traditions could be reflected in recent voting trajectories.
The abovementioned empirical research on the topic presents at least two draw-
backs: first, almost all the studies aiming to explain the new Legas geographical
trajectories of success with the scattered neo-fascist legacy of the context are mainly ag-
gregate analyses, based on official electoral data at the municipality or provincial level,
over a small part of the territory (Mancosu, 2015; Mancosu and Ladini, 2018). This ap-
proach, which in addition does not take into account southern regions that saw a massive
increase in support for the Lega in 2019, keeps open concerns regarding the inference to
the entire part of the country exposed to the partys outbreak and exposes these studies to
issues of ecological fallacy (Robinson, 1950). In other words, just by looking at the aggre-
gate, we do not have the smoking gunof the effect of the context on individual attitudes
and behaviours.
2.3. Hypotheses
The aim of this paper is twofold. First, we assess whether contextual characteristics are
able to explain, by means of an aggregate measure, the variance of likelihood to vote for
the Lega explicitly at the individual level, net of several other individual-level controls.
Second, we extend previous analyses (focusing on a few regions) to the entire country, by
looking in particular at central and southern regions, in which the Lega was an (almost)
irrelevant competitor before Salvinis leadership, and in which the increase of consen-
sus has been more impressive in relative terms. These regions also made up the most
important basin of the MSI during the First Republic (see Ignazi, 1998), leading to a more
precise test of our theoretical framework. In addition, while the diffusion of Lega Nord
in northern regions took place in a period in which the party had ethno-regionalist con-
notations, the unprecedented diffusion of the Lega beyond its strongholds coincided
with the shift of the party to nationalistic and more radical right-wing positions. In this
respect, we hypothesize that the effect of MSI strength on individual support for the Lega
will be stronger in central and southern regions. The first hypothesis will thus read as
H1. Contextual levels of the neo-fascist minority are positively associated with in-
dividual propensities to vote for the Lega in the 2019 European elections, especially
in the centre-south.
As an additional drawback, previous studies do not take into account the transfor-
mation of the extreme right-wing parties in the past decades. A correlation between the
strength of the Lega (either at the aggregate or individual level) in recent elections and
the permeability of the context to the MSI, per se, does not automatically lead to a corrob-
oration of the mechanism depicted above. As stressed in the literature, indeed, the neo-
fascist right has undergone several vicissitudes since the end of the First and the begin-
ning of the Second Republic (see Ignazi, 1998). In 1993, the MSI leadership abandoned
the ideological references to fascism (which clearly characterized the MSI during the
First Republic), by dissolving the old party and founding Alleanza Nazionale, a more
moderate party that aimed at qualifying itself as a potential governing party. The ideo-
logical shift, similar to the one that invested the centre-left parties in the same period,
allowed this new party to receive 5 million votes at its first appearance in the 1994 elec-
tions, almost doubling the historical best performance of the MSI (almost 3 million votes
in 1972). However, the relationship between the two parties remained quite strong. For
instance, the municipal-level correlation between the Movimento Sociale Italiano in
1976 and Alleanza Nazionale in 2001 two periods in which the voting percentage for
the parties was about one percentage point higher than their average is indeed equal to
.54 (authorsanalysis). It is thus possible that the places in which the MSI was strong
became progressively less extreme, aided, in this process, by the decreasing extremism
of the political entrepreneur which they referred to. The effect of the neo-fascist tradi-
tion on the Lega must thus be discounted by the effect of a subsequent, and less extreme
political entrepreneur that might have mitigated the original extreme ideological tenet
of MSI in the context. The second hypothesis will thus read as follows:
H2. Contextual levels of the neo-fascist minority in the centre-south are positively
associated with individual propensities to vote for the Lega in the 2019 European
elections, irrespective of the contextual performance of Alleanza Nazionale during
the Second Republic.
3. Data and methods
To test our hypotheses, we employed a survey carried out in Italy before the 2019 Euro-
pean elections (from 13 March to 15 May, 2019). Overall, 11,063 respondents from an
opt-in panel of a private company (SWG) were interviewed through the CAWI (Com-
puter Assisted Web Interviewing) mode. The survey contains questions referring to the
socio-demographic, attitudinal, and behavioural characteristics of the respondents. In
addition, the survey contains information on the municipality in which the respondents
live. This piece of information is employed here to link the individual dataset with the
election results extracted from the official database of the Italian Ministry of the Inte-
rior. In particular, the voting percentages for the Movimento Sociale Italiano in 1976
The distribution of the voting percentages for the MSI in the 1976 national elections by geographical area
are reported in Appendix 1. To take into account possible bias due to the extreme values of the MSI
strength in some municipalities, we also attempted to fit the same models without these extreme values.
Results, available on request, are substantially equal to those presented in this article.
The neo-fascist territorial legacy and the success of the Lega
and Alleanza Nazionale in 2001 at the municipality level have been matched with every
respondent declaring a non-missing municipality of residence.
In this paper, we present four different models with the aim of providing consistent
empirical evidence to back our hypotheses. The dependent variable is the individual pro-
pensity to vote for the Lega, measured by means of a 0-10 scale in which the respondents
express the likelihood of their voting for the party, with 0 meaning 0% likelihoodand
10 meaning 100% likelihood(for more information on the measure see van der Eijk et
al., 2006). The first model, fitted only on respondents coming from the northern regions
(northeast and northwest), explains the variation of the propensity to vote for the Lega
conditional to the level of support for the MSI at the municipal level in 1976. The second
model also includes the share of votes for Alleanza Nazionale at the municipal level as
control variable. In this way, we are able to assess whether the propensities to vote for the
Lega are associated with the permeability to the neo-fascist tradition of the context in
which people live, controlling for the level of the (more moderate) heir to this tradition
in a subsequent election (Alleanza Nazionale). The same two models are fitted only on
the respondents living in central and southern regions (third and fourth model)
. All the
models also include a set of individual-level control variables: gender, age group (six cat-
egories: <25, 25-34, 35-44, 45-54, 55-64, 65 and more), educational level (three categories:
primary, secondary, and tertiary), perceived wealth (five categories: wealthy, living eas-
ily, living with difficulties, living with huge difficulties, poor), and left-right self-
placement on a 0-10 scale recoded in 6 categories (left: 0-1; centre-left: 2-3; centre: 4-6;
centre-right: 7-8; right: 9-10; refused to locate).
According to our hypotheses, we could
expect that the coefficient of MSI will be positive and significant in the centre-south of
the country, net of control variables.
To control for the possible inflation of standard errors because of the hierarchical
structure of the data, we employ linear multilevel random-intercept regression models
(Snijders and Boskers, 1999). This multilevel approach has some antecedents in the
study of Italian electoral behaviour aimed at analysing the role of territorial political tra-
ditions on vote choice (Vezzoni, 2008). To further take into account the context at a
lower level of specification, all the models are fitted with a fixed-effect control that indi-
cates the Istat geographical area to which the respondent belongs.
4. Results
Table 2 shows the coefficients for the four multilevel regression models, fitted on re-
spondents living in the northern (Model 1 and 2) and central-southern (Model 3 and 4)
To stress the differences in voting behaviour for the League between north and centre-south we em-
ployed a separated model approach instead of estimating a single pooled model with an interaction
between MSI strength and a dichotomous north/centre-south variable. This choice is explained by both
theoretical and methodological reasons: first, insofar as we deal with the geographically scattered support
for the Lega, subdividing north and centre-south signals that we are dealing with two different subna-
tional contexts that should be treated in different ways (see Passarelli and Tuorto, 2012). Second, by
applying a two-model approach, we provide a more conservative interpretation of the effects that, in this
way are estimated separately in the two subnational contexts (a pooled approach would indeed provide a
weighted average of the control variables’ effects without taking into account geographical specificities).
All the question wordings can be found in Appendix 2.
With regard to the control variables, although we can detect some differences be-
tween the models, results are in line with previous literature (Passarelli and Tuorto,
2018). All over the country, the propensities to vote for the Lega are far higher among
respondents with a low educational level, whose average value is more than one point
higher than tertiary educated ones, both in northern and central-southern regions.
Table 2. Multilevel linear regression models to study the propensity to vote for the Lega
Model 1
Model 2
Model 3
Model 4
North only
North only
Centre-south only
Centre-south only
Indep. Variables
Lega PTV
Lega PTV
Lega PTV
Lega PTV
MSI municipal level
AN municipal level
Gender: Female
Age: 25-34 (ref. <25)
Education level: Sec-
ondary (ref. Primary)
Income: I can live
easily (ref. Im wealthy)
I have difficulties
I have huge difficulties
I feel poor
Left-right s-p: Centre-
right (ref. Right)
Not located
Level-2 variance
Level-1 variance
Number of groups
Notes: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 (the models contain fixed effect for the geographical area
coefficients not shown).
When looking at gender, women present on average a propensity to vote for the Lega
which is 0.3 higher than men. Concerning age groups, the analysis shows that, in the
south, older voters (age > 64) have the lowest propensity to vote for the Lega, In the north,
The neo-fascist territorial legacy and the success of the Lega
45-54 year old individuals are more likely to choose the Lega than the youngest ones. As
expected, individuals who locate themselves at the extreme right pole of the left-right
scale had the highest propensity to vote for the Lega in the 2019 European elections. In
line with the argument on the radicalization of the Lega, they are even significantly more
supportive of the party compared to respondents locating themselves on the centre-
right. Individual economic conditions seem to be associated with a preference for the
Lega only in the north, with people with difficulties having a higher propensity to vote
for the party.
Figure 1. Predicted propensities to vote for the Lega according to the different levels of MSI at the mu-
nicipal level in 1976 (95% confidence intervals) centre-south only, Model 4 predictions
Our two substantively relevant independent variables are, as stressed above, the
municipal-level strength of the MSI in 1976, and the level of the less extreme heir to this
party, Alleanza Nazionale, about twenty years before the 2019 elections. Results seem to
support our hypotheses. The association between the strength of the MSI in the munici-
pality and the propensity to vote for the Lega in 2019 is positive and significant, but only
in the south. In the north, the traditional stronghold of the party, we find no correlation
between MSI performance and individual support for the Lega. This is further proof of
the fact that the neo-fascist explanation of party trajectories can be an explanation only
in the centre-south of the country.
Another relevant result is that this outcome is stable even when controlling for the
municipal levels of Alleanza Nazionale, the less extreme heir to the MSI, which turns out
to be non-significant in explaining the variation in the propensity to vote for the Lega.
Since the two variables might present issues of multicollinearity, we performed a variance inflation fac-
tor (VIF) test on the model. In both north and centre-south models, the VIF score of MSI and AN variables
are around 2, a value that reassures us regarding the absence of multicollinearity issues (in general, a VIF
value over 5 or 10 starts to be worrisome; see Hair et al., 2010).
An analysis of the magnitude of the effects is enlightening as to the relevance of the
MSI effect in shaping votersattitudes and behaviours. As figure 1 shows, if we select the
central 90% of the distribution of our independent variable, the predicted propensity to
vote for the Lega (on a 0-10 scale) is equal to 2.8 when people live in a municipality in
which the share of votes for the MSI was around 2%. Instead, when people exposed to ar-
eas which are more permeable to the neo-fascist minority (level of MSI in 1976 equal to
14% on valid votes), their average propensity to vote is 0.8 points higher, equal to 3.6.
5. Conclusion and discussion
This paper aims to test the relevance of one possible explanation for the success of the
Lega in the 2019 European elections in Italy, by focusing especially on central and south-
ern regions where the consensus for the Lega was irrelevant before Matteo Salvinis
leadership. To do so, we moved from the previous literature (see Mancosu, 2015; Man-
cosu and Ladini, 2018) by looking at the persistence of the neo-fascist political tradition
in the vote for the new Lega. According to this view, one of the communication strategies
of Salvini started from the standpoint that support for the Lega could expand by exploit-
ing a potential national basin of extreme right-wing voters who were, if not contiguous,
not reluctant to a classical set of neo-fascist claims (which include Ventennio nostalgia,
a supremacist view of the Italian cultural and ethnic milieux, and a general preference
for a strong leaderfigure). Some of the rhetoric in the communication strategy of the
Legas leader followed this general trend and progressively shifted its claims and politi-
cal action to the extreme right of the political spectrum and to the centre-south as
regards the geographical diffusion trajectories. Previous research aimed at finding an as-
sociation between the geographical scattering of the Lega in the last 4-5 years and the
contexts in which the extreme right minority was stronger, were mainly based on aggre-
gate data a strategy that does not allow us to clearly identify causal mechanisms.
In this paper, we aimed to explain the electoral success of the Lega in 2019 by ex-
plaining the individual variation of propensities to vote with levels of the neo-fascist
minority in the First Republic measured by means of the municipal strength of the Mo-
vimento Sociale Italiano. The mechanism hypothesized here can be identified with a
form of behavioural path dependence (Acharya et al., 2018) that argues that contexts
more open to accepting a strong neo-fascist minority are those in which more extreme
ideas can spread, even if more than forty years have passed.
The analyses presented here provide empirical evidence towards our hypotheses, by
showing a significant association between support for the MSI and the propensity to vote
for the new Lega, but only in those areas where the Lega was almost absent before Sal-
vinis leadership. We think that our results can be relevant in showing the persistence of
political attitudes over time. Similarly to other cases of behavioural path dependence
shown above, the attitudes of the neo-fascist minority at the local level which were pre-
sent, and sometimes relevant, during the whole of the First Republic, show up again with
the exploits and communication shift of the new Lega, after a period in which they have
been hidden.
These results must be evaluated by considering at least two caveats. The first one is
that results do not show that Lega voters are neo-fascists, nor that neo-fascists mas-
sively vote for the Lega; rather, what we show here is that people in contexts in which a
The neo-fascist territorial legacy and the success of the Lega
neo-fascist minority were particularly present might have been socialized in a political
context in which, among other things, shreds of the ideology were present. It is not nec-
essary for people voting nowadays for the Lega to accept the complete set of ideological
tenets of neo-fascism; nor do they need to completely understand all the consequences
of the ideological structure, which, by the way, has been historically extremely vague and
foggy. In addition, the relationship between the MSIs prevalence and the propensity to
vote for the Lega allows us to explain only a part of the success of the party in the centre-
south, which can also be accounted for by more recent trends, such as the immigration
and economic crises (see Albertazzi et al., 2018).
Finally, our results only indirectly argue that the trigger of the correlation that we
see is the shift of Salvinis communication. Although previous studies stressed that the
correlation between the neo-fascist legacy and the trajectories of diffusion of the Lega
are related only after the advent of Salvini as leader of the party, we do not present any
evidence relating to this (crucial) point. Future research should aim at better identifying
the path dependence mechanism by means, for instance, of panel data.
Acharya, A., Blackwell, M., & Sen, M. (2018). Deep roots: How slavery still shapes southern pol-
itics. Princeton: Princeton University Press.
Albertazzi, D. (2016). Going, going,… not quite gone yet? Bossis Legaand the survival of the
mass party. Contemporary Italian Politics, 8(2), 115-130.
Albertazzi, D., Giovannini, A., & Seddone, A. (2018). No regionalism please, we are Leghisti!
The transformation of the Italian Lega Nord under the leadership of Matteo Salvini. Re-
gional & Federal Studies, 28(5), 645-671.
Bellucci, P. & Segatti, P. (Ed.). (2010). Votare in Italia: 1968-2008: dallappartenenza alla
scelta. Bologna: Il Mulino.
Bracciale, R., & Martella, A. (2017). Define the populist political communication style: the case
of Italian political leaders on Twitter. Information, Communication & Society, 20(9),
Brunazzo, M., & Della Sala, V. (2016). Italy between ‘trasformismo’ and Transformation. In Ma-
gone, J. M., Laffan, B., & Schweiger, C. (eds.). Core-periphery relations in the European
Union: Power and conflict in a dualist political economy. London: Routledge, 216-227.
Brunazzo, M., & Gilbert, M. (2017). Insurgents against Brussels: Euroscepticism and the right-
wing populist turn of the Lega Nord since 2013. Journal of Modern Italian Studies, 22(5),
Castelli Gattinara, P. (2018). Neo-fascist movement parties in Italy: The extreme right be-
tween electoral and protest politics. In Caiani, M. and Císař, O. (eds.). Radical Right
Movement Parties in Europe. London: Routledge.
De Giorgi, E., & Tronconi, F. (2018). The center-right in a search for unity and the re-emer-
gence of the neo-fascist right. Contemporary Italian Politics, 10(4), 330-345.
Diamanti, I. (2003). Bianco, rosso, verde... e azzurro. Mappe e colori dellItalia politica. Bolo-
gna: il Mulino.
Galli, G., ed. (1968). Il comportamento elettorale in Italia: Unindagine ecologica sulle elezioni
in Italia tra il 1946 e il 1963. Bologna: il Mulino.
Hair, J., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th
ed.). Upper saddle River, New Jersey: Pearson Education International.
Haney-López, I. (2014). Dog whistle politics: How coded racial appeals have reinvented racism
and wrecked the middle class. New York: Oxford University Press.
Ignazi, P. (1998) Il polo escluso. profilo storico del Movimento sociale italiano. Bologna: il Mulino.
Mancosu, M. (2015). La pista nera. Il successo della Lega in Toscana e leredità del Msi. In Pa-
paro, A., & Cataldi, M. (Eds.), Dopo la luna di miele. Le elezioni comunale e regionali fra
autunno e primavera. Roma: CISE.
Mancosu, M., & Ladini, R. (2018). The newLeague success in the red belt and its post-fascist
inheritance: evidence from 2018 National Elections. ArXiv. DOI: 10.31235/
Passarelli, G., & Tuorto, D. (2012). The Lega Nord goes south: The electoral advance in Emilia
Romagna: A new territorial model?. Political Geography, 31(7), 419-428.
Passarelli, G., & Tuorto, D. (2018). La Lega di Salvini: estrema destra di governo. Bologna. il
Rame, S. (2018). Salvini omaggia i monumenti di Roma: Non accusatemi di nostalgie musso-
liniane. Il Giornale. 2018-10-11.
Robinson W. S. (1950). Ecological correlations and the behavior of individuals. International
Journal of Epidemiology 38(2), 337-341.
Segatti, P., & Vezzoni, C. (2011). Alla ricerca del limes padano. Limes. 2, 59-70.
Serricchio, F. (2018), Il peso dellEuropa nel voto 2018. In Itanes (ed.), Vox populi. Il voto ad
alta voce del 2018. Bologna: il Mulino, 165-178.
Sniderman, P. M., Peri, P., de Figueiredo Jr, R. J., & Piazza, T. (2002). The outsider: Prejudice
and politics in Italy. Princeton: Princeton University Press.
Snijders, T. A. S. R. J., & Bosker, R. J. (1999). Multilevel analysis: an introduction to basic and
advanced multilevel modeling. London: Sage.
Tarchi, M. (1998). Italy: the Northern League. In de Winter, L., & Türsan, H. (Eds.), Regional
Parties in Western Europe. London: Routledge, 143-157.
Van der Eijk, C., Van der Brug, W., Kroh, M., & Franklin, M. (2006). Rethinking the depend-
ent variable in voting behavior: On the measurement and analysis of electoral
utilities. Electoral Studies, 25(3), 424-447.
Vezzoni, C. (2008). Territorial context and voting behaviour in the 2006 elections: A multi-
level approach. Polis, 22(2), 193-220.
Vezzoni, C. (2018). Immigrazione e insicurezza economica nelle urne. In Itanes (ed.) Vox po-
puli. Il voto ad alta voce del 2018. Bologna: il Mulino, 147-163.
Voigtländer, N., Voth H-J. (2012) Persecution Perpetuated: The Medieval Origins of Anti-Se-
mitic Violence in Nazi Germany.Quarterly Journal of Economics 127 (3): 13391392.
Wittenberg, J. (2006) Crucibles of Political Loyalty: Church Institutions and Electoral Conti-
nuity in Hungary. Cambridge, UK: Cambridge University Press.
The neo-fascist territorial legacy and the success of the Lega
Appendix 1. Voting percentages for the Movimento Sociale Italiano in 1976 national elections
Geographical area
MSI 1976
Note: Within the centre, in Lazio the only region of the macroarea not belonging to the so called ‘red zone’ the voting per-
centage for the MSI was equal to 9.5.
Appendix 2. Question wording (authors’ translation)
Dependent variable
Propensity to vote for the Lega: What is the likelihood that you will vote for the Lega in the
upcoming European elections?
1. not at all likely11. totally likely
(variable rescaled on a 0-10 scale in the analyses)
Independent variables
Gender: you are...
1. Male
2. Female
Age: Age class at 6 pre-coded in the dataset.
1. <25
2. 25-34
3. 35-44
4. 45-54
5. 55-64
6. >64
Education level pre-coded in the dataset
1. Primary
2. Secondary
3. Tertiary
Your income allows you to live...
1. Im wealthy
2. I can live easily
3. I have difficulties
4. I have huge difficulties
5. I feel poor
Politically, you would locate yourself more on the...pre-coded in the dataset
1. Right
2. Centre-right
3. Centre
4. Centre-left
5. Left
6. I would not locate myself
Geographical area pre-coded in the dataset
1. Northwest
2. Northeast
3. Centre
4. South
5. Islands
In 2018 Italian general elections, the Lega has dramatically increased its consensus in the ‘red belt’, the central area of the country traditionally ruled by centre-left parties. Pundits have argued that this performance can be attributed to the new leadership of Matteo Salvini, who changed the ideological location of the Lega by transforming it in a national right-wing party. This article hypothesizes that geographical trajectories of diffusion of the party are associated with the prevalence of a neo-fascist minority present in the ‘red belt’ during the First Republic. By employing official electoral results at the municipality level, the article analyses the relationship between the percentages of votes for the MSI in 1976 and the Lega in the 2006–2018 period in Tuscany, Umbria, and Marche. The neo-fascist inheritance results to be associated with the vote for the Lega only in 2018, after the change in the political discourse and leadership of the party.
Full-text available
In 2018 national elections, the Lega, an Italian xenophobic right-wing party, has dramatically increased its consensus in the ‘red belt’, the central part of the country traditionally ruled by center-left parties. Pundits have argued that this performance can be attributed to the effect of the new leadership of Matteo Salvini, who shifted the ideological location of the party (that now aims at being a national right-wing party), combined with the drop in preferences of Forza Italia, the ally/competitor in the right-wing ideological spectrum. This paper aims at providing new insights in the explanation of these electoral outcomes, by hypothesizing that geographical trajectories of diffusion of the party are correlated with the presence of geographically clustered post-fascist minorities present in the region since the First Republic age. By employing official figures at the municipality level, the paper analyses the relationship between the percentages of votes for the MSI (the most relevant post-fascist force during the First Republic) in 1976 and the Lega Nord in the 2006-2018 period. Consistent with our hypothesis, the post-fascist inheritance is significantly correlated with the local prevalence for the Lega Nord in 2018, after the change in the political discourse and leadership of the party. Empirical analyses provide evidence of our expectations, even when controlling for unemployment rate and percentage of immigrants.
Full-text available
Whilst the Lega Nord has traditionally been defined as a regionalist populist party, since Matteo Salvini became its leader in 2013 it has undergone a process of profound ideological transformation. This article assesses this momentous change and the impact it could have on the future of the Lega, drawing on a content analysis of Salvini’s and the party’s Facebook posts, as well as interviews with regional leaders. It argues that, under Salvini’s personal style of leadership: (a) regionalism has been replaced by an empty form of nativist nationalism, which fails to address socio-economic issues related to the North–South divide; (b) populism remains central to the party’s strategic communication, but the EU has taken Rome’s place as the people’s ‘enemy’; (c) this ideological shift has paid-off at the 2018 general election, but is underpinned by latent fractures between the leader and regional representatives which could have profound implications in the future.
In 2017, the parties of the centre-right camp – Silvio Berlusconi’s Forza Italia, Matteo Salvini’s (Northern) League and Giorgia Meloni’s Fratelli d’Italia – faced the puzzle of deciding whether to participate in the imminent general elections as allies or as rivals. On the one hand they had partially different aims (especially in relation to the European Union issue) and the problem of choosing the leadership of the alliance; on the other, the new electoral system, partially based on single-member districts, forced them to present common candidates in order to maximize their chances of reaching a majority of seats in parliament. In this article, we outline the stages through which the centre-right parties attempted to solve this puzzle over the course of the year, finally reaching an agreement to make a formal electoral alliance. In the final section of the article, we focus on one additional political actor of the right wing: the neo-fascist movements, notably Forza Nuova and CasaPound, which received significant media coverage during 2017. We describe how these movements were able to gain center stage and what this implied for the mainstream centre-right parties.
Under the leadership of Matteo Salvini, the Lega Nord has shifted away from its previous political identity as a voice for Italy’s north and has placed hostility towards the policies and institutions of the European Union (EU) at the heart of its rhetoric. Nowadays, the enemy is Rome no longer: it is Brussels, European institutions, and the threat to the national sovereignty posed by the EU. Borrowing from the Italian political philosopher Nicola Matteucci, we would describe Salvini’s Lega as a ‘populist insurgency’. That is to say, it is a populist party that marries the traditional populist evocation of the virtues of the people against the corrupt elites, with a pervasive glibness of analysis.
In the hybrid media system, many processes are reforming political communication: popularisation, disintermediation, personalisation, intimisation and of course populism. This study proposes an empirical definition of political communication style with the aim of identifying characteristics of the populist political communication style. Between 2015 and 2016, the Twitter timelines of the main political leaders in Italy were analysed for 16 months. Applying an MCA allowed us to identify two key factors that characterise the communication styles of leaders: (1) communication mode, comparing negative and positive; and (2) communicative focus, comparing personalisation and political/campaign. The intersection of these two factors resulted in four different political communication styles: ‘Engaging’, ‘Intimate’, ‘Champion of the people’ and ‘Man of the street’. The latter two were clearly characterised by the presence of populist ideology fragments and traits, but were not strictly related to the leaders’ ideological positions. This result supports the hypothesis that populist style is less and less connected to the right/left political cleavage, but rather the result of a varied combination of gradations that mix different individual aspects of the leader’s political communication style.
This book investigates one of the oldest paradoxes in political science: why do mass political loyalties persist even amid prolonged social upheaval and disruptive economic development. Drawing on extensive archival research and an original database of election results, this book explores the paradox of political persistence by examining Hungary's often tortuous path from pre- to post-communism. Wittenberg reframes the theoretical debate, and then demonstrates how despite the many depredations of communism, the Roman Catholic and Calvinist Churches transmitted loyalties to parties of the Right. Contrary to conventional wisdom, Church resistance occurred not from above, but from below. Hemmed in and harassed by communist party cadres, parish priests and pastors employed a variety of ingenious tactics to ensure the continued survival of local church institutions. These institutions insulated their adherents from pressures to assimilate into the surrounding socialist milieu. Ultimately this led to political continuity between pre-and post-communism.