Framing the context: the strengthening of the meso-level in Italy

Throughout the 1990s, Italy implemented a series of reforms to increase regional autonomy to move beyond the high level of centralization and the national government's undisputed dominance over the meso-level (Baldi, Reference Baldi2020). The 2001 Constitutional Reform (Riforma del Titolo V), which provided regional governments exclusive competencies over a wide range of policy sectors and expanded legislative powers, marked the pinnacle of this non-linear devolution process (Salvati, Reference Salvati2022).

Despite the impressive push towards decentralization, adequate institutional tools to ensure coordination have yet to be planned or deployed, allowing constant tensions and intergovernmental conflicts among the various levels of government to flourish (Baldini and Baldi, Reference Baldini and Baldi2014; Bolgherini, Reference Bolgherini2014). To further complicate matters, another major institutional reform was implemented in 1999, establishing direct elections of regional presidents and allowing each region to determine its statute and electoral law. Over the years, regional presidencies in Italy have constantly strengthened power and resources through a concentration of political and administrative functions in the presidency (Musella, Reference Musella2009; Passarelli, Reference Passarelli and Vassallo2013). This reform has elevated the regional executive to a supraordinate position at the expense of the legislative assembly, fuelling the presidentialization of Italian regions (Musella, Reference Musella2009; Vampa, Reference Vampa2015).

This pattern of bolstering executive posts at municipal and regional levels is common in Europe (Guérin and Kerrouche, Reference Guérin and Kerrouche2008; Stolz and Fischer, Reference Stolz and Fischer2014; Astudillo and Martínez-Cantó, Reference Astudillo and Martínez-Cantó2020). In terms of political career patterns, such a movement of decentralization in favour of meso-level institutions has made available more political positions with more authority and resources, making them more appealing to professional politicians (Di Capua et al., Reference Di Capua, Pilotti, Mach and Lasseb2022). A regional political career is no longer a simple springboard for national political venues in a unidirectional conception of career but a more interesting option per se for politicians (Tronconi, Reference Tronconi, Best and Higley2018), implying a more complex and stratified model of elite circulation. Consequently, this calls to enquire how such dynamics have transformed the profile of political professionals at the local level. In Spain, UK and France, the background of regional chief executives is growingly characterized by the rising importance of previous experience in regional parliament/government – leading to ‘regionalization’ of the political class – and the contextual declining relevance of prior political experience at the national level (Botella et al., Reference Botella, Rodríguez Teruel, Barberà and Barrio2010). The climb to the apical position of the regional executive seems to be consequential to the unfolding of a successful career at the regional level, framing for these countries an alternative model of political career (Borchert and Stolz, Reference Borchert and Stolz2011). This pattern is also confirmed by evidence from Germany, which revealed that the gain of a cabinet position at the regional level is mainly the result of a pure regional career, with the cabinet office that is often directly related to having had a regional parliamentary experience before being appointed as regional minister (Stolz and Fischer, Reference Stolz and Fischer2014). In Spain, strengthening the regional institutions has acted as a powerful catalyst for professional party insiders (Astudillo and Martínez-Cantó, Reference Astudillo and Martínez-Cantó2020).

By contrast, in Italy, the concurrent decentralization and presidentialization processes have led to an increase in the number of chief executives with no previous political experience (outsiders) or with a national political background instead of a solid subnational political experience, with the partial exception of mayors of capital municipalities (Grimaldi and Vercesi, Reference Grimaldi and Vercesi2018). The growing attractiveness of the leadership positions at the regional level, particularly the position of chief executive, has set the stage for an open competition among politicians with extremely variegated backgrounds.

Before the reform, the presidents were elected by regional parliaments, and they were strictly dependent on the confidence of a political majority within the chamber, which was in charge of defining the regional governments' policies and their implementation, resembling the model of parliamentary-party government operating at the national level (Musella, Reference Musella2009). Moreover, the regional cabinet members were party members accountable more to the party than to the president for their political actions (Passarelli, Reference Passarelli and Vassallo2013). Indeed, regional governments, like their national counterparts, were subject to frequent political crises that were used to produce a change in the regional president and/or in the political majority. The 1999 reform changed almost everything because: ‘Regional presidents now nominate regional ministers directly and can hire and fire them. They do not have to be chosen from among the members of the regional parliament as was the case in the past, so the president is free to autonomously select ministers outside the rank and file of parties and/or the Council, determining that they are politically dependent from the president trust. Presidents are free to establish and steer the executive's policy by means of a strong team of staff (Salvati, Reference Salvati2022: 8)’.

Regional ministers are the apex of regional administrative offices and are in charge of ruling over sectors on which the regional governments have specific authority. Anyway, given that the regional statutes affirm the principle of collegiality of the Giunta (the regional cabinet) in policy-making activity, the regional ministers are called upon to implement the collective body's choices in the areas of their competence based on the portfolios' distribution made by the president. This means that regional ministers have minimal policy-making autonomy. In any case, it is impossible not to notice how, over time, the competencies granted to regional ministers, who sometimes have significant powers over sectors of primary importance for regional governments, have frequently translated into the conferral of the power to determine the substantial content of policy decisions (Fenucci, Reference Fenucci2009).

Overall, however, it is worth noting that regional elections are now more individualized and have shifted from party-centred to candidate-centred, with political parties being marginalized, as seen by their lack of authority over regional policy-making (Vampa, Reference Vampa2021). The electoral system for regional elections is mixed and consists of the concurrent election of the regional president with plurality voting and the council with a proportional system with a majority bonus. The bonus ensures that the president has a solid majority in the council. However, the high level of differentiation (Vampa, Reference Vampa2015) between regions resulted in a cross-regional fragmentation of the details of the electoral law, which are all proportional but with different thresholds, district magnitude, majority bonuses and gender quotas.

Structure of the dataset

The ITREGPARL dataset, collecting and providing information about Italian regional councils between 1993 and 2020, offers a new tool to study Italian regions. The dataset is open access and freely available on the online repository Italian Political Science Review (IPSR/RISP) Dataverse of Harvard Dataverse, together with our coding scheme.Footnote ³ The primary data source is the Anagrafe Amministratori, released by the Italian Ministry of Interns and available online.Footnote ⁴ We collapsed the available information by legislative term, recording the names, roles and personal information of the members of the regional councils. We thus cross-checked the information and missing data with ‘Openpolis’Footnote ⁵ and other online sources such as official regions' websites. We then coded and transformed the raw information to make it available as data.

Our dataset comprehends five regional legislatures – from the 6th to the 10th – for the 20 Italian regionsFootnote ⁶ and 47 variables. The unit of observation is the politician-legislature, for a total of 6077 observations. To account for politicians elected multiple times, each individual is assigned a unique ID, for a total of 4078 politicians. Thus, for each legislative term and region, we recorded members' roles – whether they are regional councillors, regional ministers, regional presidents or vice-presidents, regional council presidents and vice-presidents – gender, age, former profession, level of education, previous experience, number of mandates, length of office, incumbency, party, party ideology,Footnote ⁷ party ideological group, non-partisan lists (civic lists). For regional ministers, we recorded whether they are elective or non-elective. We also included region-level variables: region name, geographical area, NUTS (Nomenclature of territorial units for statistics) macro-region,Footnote ⁸ presence and type of gender quota,Footnote ⁹ RAIFootnote ¹⁰ and type of statute.Footnote ¹¹ As such, our dataset covers 5055 regional councillors, 1174 regional ministers, 184 vice-presidents and 118 presidents of the regional council, 79 vice-presidents and 117 regional presidents.Footnote ¹² Table 1 presents an excerpt of the ITREGPARL dataset. The full list of 47 variables is presented in the online codebook.Footnote ¹³

Table 1. Excerpt of the ITREGPARL dataset

Potential applications

The ITREGPARL dataset can be easily applied to investigate variations in the composition of the regional political class and to test hypotheses related to specific factors leading to a differentiation of regional politicians' profiles. For example, it allows replying to a series of questions related to the existence of a feminization process (Lovenduski, Reference Lovenduski2005; Beauvallet and Michon, Reference Beauvallet and Michon2010; Kenny, Reference Kenny2013; Kenny and Verge, Reference Kenny and Verge2013) in Italian regions by investigating the share of women in regional councils and the presence of different profiles for male and female councillors. Figure 1 summarizes some descriptive statistics about the gender composition of regional councils and the characteristics of male and female councillors.

Figure 1. Feminization of Italian regional councils and the profiles of female and male councillors (percentage values).

Firstly, our data show that female representation in the regional councils has constantly increased since the seventh legislative term. Nevertheless, the share of women in regional parliaments only attests to 17.35% in the 10th legislature, the most recent of the terms observed. The low degree of feminization of regional councils is even more striking compared to the share of female representatives in the Italian national parliament, which shows rates of female members of parliament above 30% since 2013. Overall, accordingly, at the regional level, gender equality is not pursued to the same degree.

Secondly, our data allow us to investigate further the personal characteristics of the members of regional parliaments. Accordingly, we observe that female councillors are overall younger and more educated than their male counterparts consistently over time, thus confirming the findings of the international literature (see Escobar-Lemmon and Taylor-Robinson, Reference Escobar-Lemmon and Taylor-Robinson2009; Stockemer and Sundström, Reference Stockemer and Sundström2019).

Lastly, apparent gender-based differences emerge by enquiring into the experience in the office of regional councillors. The share of male councillors with at least one mandate in the period under study is constantly higher than their female counterparts. However, it is interesting to observe some convergence in the amount of experience in the regional council of male and female councillors in the 10th legislature. However, the convergence is not given by an increase in the share of female councillors with at least one previous mandate but, by contrast, by a sheer decrease in the share of experienced male councillors.

Further research on the feminization of Italian regional politics may rely on the ITREGPARL dataset, for example, to assess the differences in women's presence between regional councils and regional governments, to examine regional variations, or to investigate the determinants of such processes.

A second potential application of the ITREGPARL dataset moves away from the study of politicians' characteristics and points to the investigation of the effects of the presidentialization process on the structure of regional electoral competition. The strengthening of the regional president vis-à-vis the regional council (Salvati, Reference Salvati2022) has modified political competition, leading to a tendency of the chief executive candidates to detach themselves from national parties (Bolgherini and Vampa, Reference Bolgherini, Vampa, Callanan and Loughlin2021; Vampa, Reference Vampa2021). As Vampa writes: ‘regional leaders seem increasingly to rely on a number of local or personal lists, avoiding partisan labels, to reach new voters and strengthen their positions and autonomy (Vampa, Reference Vampa2021: 179)’. The ITREGPARL dataset provides a valuable instrument to test these expectations. Specifically, it allows the investigation of longitudinal and cross-regional variations in the fragmentation of the electoral system. It can be used, for example, to assess the number and the share of seats of non-partisan electoral lists (civic lists) in regional councils.

As it emerges from Figure 2, the overall share of seats obtained by non-partisan lists as the result of regional elections has constantly increased in the period under investigation, moving from 3.97% in the 6th legislature to 18.74% in the 10th, thus confirming the effect of the presidentialization process hypothesized by the literature on the nature of regional competition. However, the most striking result is the 10 percentage points increase in seats allocated to civic lists between the 6th and the 7th legislative terms, resulting from the first regional election held after the Presidential elections reform of 1999. Accordingly, it is possible that the reform immediately affected and transformed regional electoral competition, while subsequent increases resulted from continuing adaptation to the new rules of the game.

Figure 2. Share of legislative seats of non-party lists over time (percentage values).