https://orcid.org/0000-0002-3809-3455
![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
ABSTRACT
Despite longstanding debate, research on social capital in Italy is waning. This article resumes the topic through new methodological and theoretical advances. We shed new light on the subject by combining classical social capital dimensions (civicness, social relations) with environmental attitudes. By applying an original clustering technique to individual-level data, we show that certain regions of the Mezzogiorno depart from the traditional North-South divide in social capital. The findings challenge the conventional regional cleavage by demonstrating greater pro-environment behaviours in certain Southern regions, providing fresh perspectives for sustainable development.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Supplementary material
Supplementary data for this article can be accessed at https://doi.org/10.1080/13608746.2024.2343498.
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Notes
1 A recent careful exploration of the empirical links between the bonding and bridging forms in European regions suggests that ‘while bridging social capital has a positive effect on regional economic growth when controlling for bonding, bonding social capital is negative for growth when controlling for the level of bridging in the region’ (Muringani, Fitjar & Rodríguez-Pose Citation2021, p. 1414). However, as other studies have pointed out, it is possible to detect complementary relations between the two forms of social capital. For example, Ramella and Trigilia (Citation2010) showed that innovation is enhanced when inventors are simultaneously embedded in networks with strong and weak ties. More recently, Antonietti and Boschma (Citation2021) underlined that while bridging social capital encourages new industry, it loses impact during economic crises, whereas bonding social capital enhances resilience during downturns.
2 Although an analysis of variance can provide this information, it is difficult to integrate such results into a synthetic measure.
3 Although this work addresses the issue of aggregation effects in measuring regional social capital endowment, it leaves other issues unsolved, including the discrepancy between social capital’s theoretical insights and the limitations of its empirical conceptualisation and operationalisation. For further information on this topic see (Bjørnskov & Sønderskov Citation2013; Engbers, Thompson & Slaper Citation2017; Weiler & Hinz Citation2019).
4 While this method allows us to avoid the ecological fallacy when delineating similarities and differences between regions, it is not without its shortcomings. Indeed, it exposes research results to the problem of the individualistic fallacy (Loney & Nagelkerke Citation2014; Seligson Citation2002; Subramanian et al. Citation2009). We acknowledge the intrinsic conceptual difference between variables at individual and group levels. For example, beyond personal poverty, the poverty or deprivation of the society or area in which one lives also plays a crucial role. Societal-level variables significantly impact individual ones. Thus, while our method is well-equipped to deal with compositional effects, it struggles to deal with contextual ones.
5 Principal components analysis is calculated with a Varimax rotation. While this procedure is commonly used due to its analytical convenience, it has a problem. Maximizing the independence (orthogonality) of components, while enhancing interpretability, may also lead to a loss of meaningful correlations between components, distorting their relationship. This results in two significant issues: first, it may obscure relevant correlations between variables present in the original data structure, and second, it can overshadow multicollinearity problems. However, these two issues are not particularly problematic in our research design. We do not use Varimax rotation synthetic indices as predictors but only to describe the characteristics of each cluster; thus, the multicollinearity concern is relative. Furthermore, graphic comparisons () of factor loadings allow us to explore the possible relationships between the different dimensions. Indeed, the rotations change the absolute value of factor loadings but not the relative positions of clusters among them, and consequently not the overall mean.
Figure 1. Ideal types of civicness in Italy.
Figure 2. Ideal types of relational social capital in Italy.
Figure 3. Cluster distribution in relation to environmental citizenship dimensions in Italy.
6 The decision to use exactly five clusters is based on Cubic Clustering Criterion (CCC) results, a feature of SAS statistical software. The CCC is employed to estimate the number of clusters using Ward’s minimum variance method, k-means or other approaches that aim to minimise the sum of squares within clusters. We conducted robustness tests, which confirmed the stability of the results with respect to the number of clusters. Additional evidence on this topic is provided in the Appendix (Table A2).
7 Only participation in political events revealed an ambiguous attribution of loadings. This is an unsurprising result because attending events such as political speeches highlights an attitude that is congruent with active participation and staying informed.
8 Note that cluster two is slightly above the mean for the associative and information dimensions, while cluster three shows the highest levels of both.
9 Our criterion for identifying a significant difference between regions was a threshold of ±15 per cent in cluster-relative concentrations compared to the national distribution.
10 However, Sicilia and Campania differed in terms of invisible participation, which was closer to the national level in the latter.
11 In Multi-Dimensional Scaling (MDS), a two-dimensional representation is generated to visualise similarity and dissimilarity relationships between objects or elements of the original dataset. The coordinates are calculated to reproduce the relative distances and proximities present in the original data while retaining as much information as possible. Essentially, these coordinates reduce the dimensions of the original dataset, providing a two-dimensional visual representation that preserves the inherent relationships between objects. Therefore, strictly speaking, the axes of an MDS chart lack a concrete interpretation. In the present context, we endeavoured to capture the significance of the relative positions of the 20 regions on the map.
Additional information
Notes on contributors
Nicolò Bellanca
Nicolò Bellanca is Associate professor in Applied Economics at the University of Florence, teaching in undergraduate, master, and doctorate courses. His research was published on journals in economics, political science, sociology, history and philosophy. His recent research has focused on the theory of institutional change and on local economic systems. He has authored several books and has published widely through contributed chapters and journal articles (https://www.researchgate.net/profile/Nicolo-Bellanca; https://unifi.academia.edu/Nicol%C3%B2Bellanca). He is also engaged in scientific dissemination and public debate, in magazines such as Micromega and Il Ponte.
Alberto Gherardini
Alberto Gherardini is an associate professor of Economic Sociology at the University of Torino. He received his PhD in Sociology from the University of Firenze. He specializes in research on innovation policies, university-industry relations, regional development, and employment relations. He has authored several books and articles on Italian and international journals (https://www.researchgate.net/profile/Alberto-Gherardini).
Mauro Maltagliati
Mauro Maltagliati is an associate professor of Economic Statistics at the University of Florence. His primary research interests encompass the technical analysis of efficiency, examination of the geographical distribution of the gender gap, and studies on well-being. He also delves into topics such as equivalence scales and comparisons of the standard of living among homogeneous groups of individuals.
Gianmaria L. Pessina
Gianmaria L. Pessina is post-doc fellow in economic sociology at the University of Torino (Italy), where he researched mainly on topics of local and regional development and sociology of economic innovation. His recent publications include ‘Italy at a critical juncture. Game changing crises for the innovation system’ in Stato e mercato (2022, whit F. Ramella).