Southern Italian Peripheries at the Ballot Box. Continuity and Innovation in Electoral Behaviours in Three Southern Italian Cities during Post-Crisis Decade (2008–2019)

ABSTRACT

The article explores the social and political geography of three southern Italian cities – Bari, Napoli and Reggio Calabria – and investigates inside their centre–periphery cleavage, highlighting whether and how peripheral conditions affect electoral behaviours. It provides a multidimensional statistical analysis based on I.S.T.A.T. data on Italian cities and suburbs and on the results of the political and European elections during the post-economic crisis decade (2008–2019). The study highlights a growing predictive power, after the 2008 crisis, of the peripheral condition on the electoral behaviour, the weakening of the centre-left parties and the specular political success of 5-Star Movement (M.5.S.) in deprived urban areas.

RIASSUNTO

L’articolo si propone di ricostruire la geografia sociale e politica di tre città metropolitane del Sud Italia – Bari, Napoli e Reggio Calabria – indagando al loro interno la frattura centro-periferia e, in particolare, se e in che modo la condizione di perifericità influenzi il comportamento elettorale. La ricerca si basa su un’analisi statistica multidimensionale, che prende in considerazione i dati Istat dell’indagine sulle città e sulle periferie nonché i risultati delle elezioni politiche ed europee del decennio successivo alla crisi economica (2008–2019). Lo studio mette in evidenza l’incremento del peso assunto, dopo la crisi del 2008, dalla perifericità sociale come variabile predittiva del comportamento elettorale: nelle aree urbane deprivate, i partiti di centro-sinistra si sono indeboliti a fronte di uno speculare successo politico del Movimento 5 Stelle (M5S).

KEYWORDS:

• southern Italy
• peripherality
• electoral behaviour

PAROLE CHIAVE:

• Mezzogiorno d'Italia
• perifericità
• comportamento elettorale

Notes

1. We refer to the personal movement founded by the shipowner Achille Lauro.

2. Consider to have a data set represented in terms of an matrix X consisting of n observations (neighbours, in rows) and m variables (peripherality dimensions, in column), P.C.A. transforms X to a matrix Y of same dimensions by projecting it through a orthogonal matrix P = {p_1,…, p_m} of bases Y = PX with the aim to de-correlate the original data by finding the directions in which variance is maximized and then use these directions to define the new basis. P.C.s p_i are then able to encapsulate the heterogeneity expressed by the selected indicators whilst also minimizing the loss of information (Moller 1995). The components are orthogonal and therefore uncorrelated to each other and are extracted sequentially until the total variability of the data is exhausted. The first component consists of a linear combination of data and captures the highest possible amount of variance. The second component works in the same way but acts on residual variance (unexplained by the preceding component) and so on for the remaining components. The P.C.s can be obtained by eigen decomposition of the data covariance matrix. The resulting eigenvalues are the coefficients of orthogonal transformation of variables into P.C.s or back, representing the amount of variance in the original variables accounted for by each P.C.

3. In total, sixty-one sub-municipal units were considered in the analysis: seventeen in Bari (the ‘Ex Marconi-San Girolamo-Fesca’ neighbourhood was not considered in the analysis and is included in the ‘Marconi-San Girolamo-Fesca’ neighbourhood); twenty-nine in Naples (the ‘Zona Industriale [Industrial Estate]’ and ‘Poggioreale’ neighbourhoods have been incorporated into the analysis, given the lack of details concerning the electoral variables for the two individual neighbourhoods); and 15 in Reggio Calabria.

4. The sub-municipal administrative units analysed are marked by significant heterogeneity in terms of resident population: the data varies from a minimum of 2,036 inhabitants in the former ‘Ortì-Podargoni-Terreti’ district in Reggio Calabria, to a maximum of 71,808 inhabitants in the ‘Fuorigrotta’ neighbourhood in Naples. The same is true of the neighbourhood size, which ranges from 0.42 km² for the ‘San Giuseppe’ neighbourhood in Naples, to 61.8 km² for the ‘Ortì-Podargoni-Terreti’ neighbourhood in Reggio Calabria.

5. In a first P.C.A. attempt with absolute values, the results of the Factor Analysis clearly show the influence of the scale of the individual cities on the values of some variables, particularly property prices, the centrality index and the resident foreign population percentage.

6. For the sake of brevity, hereinafter the first component shall be referred to as ‘P.C.1.’ and the second component as ‘P.C.2.’

7. Squared cosines represent the relative contribution of the factor to the explanation of each unit’s variance. When the squared cosine of a variable related to a P.C. is closer to 0, that particular variable shows a poor representation quality on the P.C. itself. In the opposite case, a squared cosine close to 1 is related with variables having good representation quality on the considered P.C.

8. The variables with PC1 squared cosine higher than 0.5 are: Incidence of Families suffering from potential economic hardship (0.92), Secondary School qualifications or Degree index of achievement (0.92), Unemployment Rate (0.91), Secondary School non-completion index (0.79), Employment Rate (0.76), Incidence of young people up to 24 years (0.75), Incidence of N.E.E.T. Young People (0.73), Average property values (0.67), Social and Material Vulnerability Index (0.66), Old-age index (0.63).

9. The variables with P.C.2. squared cosine higher than 0.5 are: Distance from city centre, journey time on Private Transport - Car (0.79), Distance from city centre (0.76), Distance from city centre, journey time on Public Transport (0.65).

Additional information

Notes on contributors

Onofrio Romano

Antonella Coco, PhD in ‘Politics, Society and Culture’, is adjunct professor in political sociology. Among others, her publications are, Reggio Calabria: le condizioni sociali in una città del Sud (Franco Angeli 2011), Politica e città in trasformazione (Rubbettino, 2018).

Antonella Coco

Lucio Palazzo, PhD in Statistical Methods at the doctoral school of Markets and Firms Economy and Policies of the University of Salerno. He is now research fellow at the University of Naples ‘Federico II’. His research mainly deals with the analysis and the development of new models for the study of historical series of territorial data.

Lucio Palazzo

Onofrio Romano is associate professor of sociology at the University of Bari ‘A. Moro’. He sits on the professors’ board of the PhD of Human Relations Sciences. Among his works are, The Sociology of Knowledge in a Time of Crisis (Routledge 2014) and Towards a Society of Degrowth (Routledge 2020).

Pietro Sabatino

Pietro Sabatino, PhD in local and regional development, is research fellow at the University of Naples ‘Federico II’. He studies the electoral behaviour in Naples and southern Italy, focusing on the local political elites and the relation between socio-economic condition and vote. Among his recent work is, Il voto del Movimento 5 Stelle nelle aree marginali. Le elezioni del 2018 a Napoli, in Meridiana n.96 (2019) (with Ciro C. De Falco).