Between export-led growth and administrative Keynesianism: Italy’s two-tiered growth regime

ABSTRACT

Comparative political economy scholarship struggles to categorise Italy’s model of capitalism between a mixed-market economy and a hybrid, stagnant economic system. To enhance our understanding of the Italian political economy, this article employs the analytical framework of growth regimes to study Italy’s regional economic systems. Our analysis indicates that Italy can hardly be defined as a ‘national growth regime’ due to the presence of two diametrically opposed regional growth regimes: Northern regions conform to a manufacturing-based, export-led growth regime supported by competitiveness-enhancing territorial institutions; southern regions conform to a particular variety of the consumption-led growth regime, that is, an administrative Keynesianism regime, which we theorise to typify a regime where growth and employment are systematically dependent on the state’s role of employer of last resort, the state’s consumption-enhancing social policies and economic forbearance of labour and corporate tax regulations. The article suggests that studying regional growth regimes is desirable when marked internal diversity in economic outcomes or productive structures exists across regions within (generally larger) countries, and when subnational governments have powers to develop major own institutions/policies in support of regional growth regimes.

KEYWORDS:

• Growth models
• varieties of capitalism
• comparative political economy
• Italy
• regional economies

Introduction

During the last decade, there has been increasing interest within the field of comparative political economy (CPE) in the study of national growth models (Baccaro and Pontusson 2016, Stockhammer and Kohler 2022) and growth regimes (Hassel and Palier 2021). Various related topics have been analysed, ranging from growth models’ varieties across countries (Bohle and Regan 2021, Baccaro and Höpner 2022, Bürgisser and Di Carlo 2023, Schedelik et al. 2021, Vukov 2023), the economic and fiscal policies underpinning growth models (Haffert and Mertens 2021), or growth model-specific patterns of inequality (Behringer and van Treeck 2022), financialisation (Ban and Helgadóttir 2022) and comparative advantages in the green transition (Driscoll 2023).

While this flourishing body of literature has contributed greatly to the field, CPE debates often overlook the spatial dimension of growth. This article problematises the predominantly national focus of the CPE growth models/regimes literature on two grounds. First, large countries are hardly ever homogenous in socio-economic terms. This applies to supply-side economic institutions (Crouch et al. 2009), which may vary in form and functions across subnational governments. And it applies to demand-side growth drivers and sectoral specialisations, which can vary substantially across regions in countries with marked internal diversity such as Italy (Trigilia and Burroni 2009) – or across cities, e.g. the City of London (Fraccaroli et al. 2023). Second, during the last three decades, most European states have undergone a process of devolution of policy competences to subnational governments (Hooghe et al. 2010). This has resulted in the greater capacity of subnational governments to enact policies and set up territorial socio-economic institutions complementing the national setting and impacting the functioning and performance of the local economy (Kazepov 2010, Scalise and Hemerijck 2022).

Thus, this article advances a regional perspective on growth models that puts emphasis on key intra-country differences in both the demand- and supply-side of regional economies. To demonstrate the contribution of our analytical approach, we leverage a case study on Italy, a classic country-case which lies uncomfortably within the usual capitalist typologies elaborated by the CPE scholarship. We employ the theoretical framework of growth regimes (Hassel and Palier 2021) and adapt it to analyse regional growth regimes’ major components, namely (1) the growth drivers of regional demand; (2) the main sectors operating as the engine of growth in the regional economy; (3) the characteristics of (sub)national supply-side institutions which shape the functioning of regional economies.

Theoretically, we build on the classic distinction between export-led and consumption-led growth regimes but introduce a novel theorisation for a specific subtype of consumption-led regime, which we theorise as administrative Keynesianism (AK henceforth) – explicitly echoing Colin Crouch’s concept of privatised Keynesianism (Crouch 2009). Crouch had defined privatised Keynesianism as an ‘unacknowledged policy regime’ in which states manage domestic demand, not through public budgets but via liberalised access to new financial instruments enhancing households’ capacity to consume out of credit. Our theorisation of AK shares similarities with Crouch’s intuition. But in the AK regime the state, through its fiscal, administrative and regulatory/enforcement powers, sustains demand, especially in backward areas where private sector alternatives remain underdeveloped. Unlike classic Keynesianism, however, AK is not simply a counter-cyclical fiscal policy for aggregate demand stabilisation. It is an unacknowledged policy regime – à la Crouch – where the process of local growth and employment generation systematically revolves around the state’s capacity to act as employer of last resort to mobilise unemployed people into the public sector, its capacity to support household incomes via consumption-enhancing social policy (Beramendi et al. 2015), and indirect support for economic activity in the shadow economy via the forbearance of labour and corporate tax regulations (Dewey and Di Carlo 2022).

Our empirical analysis indicates that, by and large, Italian regions cluster around two regional growth regimes. Northern regions (Emilia-Romagna, Friuli-Venezia Giulia, Lombardia, Piemonte, Toscana and Veneto) conform to an export-led growth regime centred on a large and resilient manufacturing sector, one that is highly integrated into international markets and endowed with strong regional supply-side institutional complementarities in support of territorial competitiveness. On the contrary, Italy’s southern regions (Calabria, Campania, Puglia, Sardegna and Sicilia) are best defined as AK regions where national consumption-enhancing social policies markedly contribute to uphold regional consumption. These regions are rather sheltered from international trade, display labour markets with relatively larger public and irregular employment and are characterised by the absence/weakness of competitiveness-enhancing territorial institutions, with the result that many firms and workers operate in the large Non-Observed Economy¹ (NOE).

This article’s contribution to CPE debates is threefold. First, by shifting the focus to the regional dimension of growth, our article points to the importance of within-country variation in both supply-side institutions and demand-side growth drivers. However, we do not claim that a regional focus is necessarily better than, or a substitute for, ‘methodological nationalism’ in the study of models of capitalism. Both approaches have merits and could be complementary. A regional approach is desirable when countries display marked internal diversity in economic outcomes/structures or when subnational governments dispose of the powers to develop subnational institutions and policies which significantly deviate from the national setting. Thus, our analytical approach could fruitfully be applied to country cases other than Italy characterised by major territorial socioeconomic disparities, e.g. Germany, Spain and the UK (Balakrishnan et al. 2022). Second, with the concept of AK, we contribute a novel growth regime typology that aims to operationalise the multifaceted channels through which the state can support consumption-led growth beyond the classic Keynesian counter-cyclical demand management. Third, we provide what is to our knowledge the first empirical application of the concept of growth models/regimes to Italy’s ‘regionalised capitalism’ (Trigilia and Burroni 2009).

The article unfolds as follows. We first situate the Italian model of capitalism within the CPE literature. Secondly, we elaborate the theoretical framework and then present our empirical results. The conclusions summarise the findings and discuss the implications of our analysis for ongoing CPE debates.

Italy’s model of capitalism between methodological nationalism and regional varieties of capitalism

The Italian model of capitalism has long fascinated scholars of political economy. Within the Varieties of Capitalism (VoC) paradigm, Italy once featured as a Mixed-Market Economy with the state playing a pivotal role in governing the economy. The state did so by shoring up the coordination capacities of fragmented and adversarial economic stakeholders (Molina and Rhodes 2007), by channelling public investment via state-owned enterprises (Celi and Guarascio 2019), and by directing credit allocation through the public ownership of large parts of the banking sector (Deeg 2009). In this sense, Italy is typified as a (weak) statist model of capitalism (Schmidt 2002), one that has been perceived as an increasingly hybrid model because the privatisations of the 1990s induced by Europeanisation decreased the state’s reach in the economy (Ferrera and Gualmini 2004) while inconsistent institutional reforms introduced elements typical of both organised and liberal economies (Simoni 2020). With the shift toward the growth models perspective in CPE (Baccaro and Pontusson 2016), Italy emerged again as a hybrid model which stagnates between weak export-led growth and feeble domestic demand (Baccaro and Bulfone 2022), and which is constrained by the Eurozone’s vincoli esterni (Celi and Guarascio 2019, Baccaro and D'Antoni 2022, Guarascio et al. 2023).

While these different perspectives have shed light on some key aspects of the Italian model of capitalism, a few shortcomings are worth discussing. First, while scholars in the VoC tradition did capture the state’s central role in the Italian economy, they were mostly interested in the state’s coordination capacities. By so doing, the VoC literature has neglected the much broader role played by the state in supporting domestic demand in Italy’s economic system. Second, in ‘re-thinking’ the VoC paradigm, the growth-model perspective has underplayed the supply-side institutions of Italy’s growth model. In this regard, Hassel and Palier (2021) have reintroduced the concept of growth regimes to combine the study of the demand-side growth drivers with supply-side institutional complementarities across regime typologies. Yet, like the VoC and growth-model perspectives, their approach continues to focus only on the national level. This overlooks the relevance of within-country variation for understanding national models of capitalism in countries such as Italy, where marked internal diversity and territorial inequality is a feature, not a bug. These are in fact aspects long highlighted in a different stream of the CPE literature which focused on the territorial dimension of the economy. Due to space constraints, we cannot do justice to this vast literature. It is sufficient to mention here a few scholarly works which, among others, have best shown the importance for CPE to analyse the local dimension of capitalism to comprehend Italy, and, more generally, countries marked by considerable internal diversity.

Among them, Richard Locke (1996, p. 484) argued that national political economies ‘are not coherent systems but rather incoherent composites of diverse subnational patterns that coexist (often uneasily) within the same national territory.’ Locke characterised Italy as a composite economy to champion a CPE approach that would pay greater attention to within country variation in institutions and patterns of state-society relations. For him, Italy’s paradox was that of a country with an incoherent economic model displaying traits of national economic decline compounded by territorial competitiveness and entrepreneurial vitality (Locke 1995). To make sense of Italy, Locke argued, one needs to look at the local ‘sociopolitical networks’ in which economic actors are embedded (Locke 1996, pp. x, xi). During the 1980s, the capacity of local economic systems to adjust to the mounting challenges of globalisation differed widely across Italy’s sclerotic national economy. Intra-country variation in patterns of economic development could only be explained by the different patterns of associationalism, interest group relations, political representation and economic governance present at the regional/local level. Along these lines, Trigilia and Burroni (2009) later challenged VoC’s methodological nationalism and suggested to shift the analysis to the regional models of capitalism to better understand countries with high internal diversity. By so doing, they spoke of Italy as a form of regionalised capitalism characterised by elements of ‘national public disorder’ (e.g. inefficient policies) and ‘a mix of local economic dynamism’ whereby regional network-based economic governance bolstered local firms’ ‘territorial competitive advantages.’

In this paper, we thus tap into both the national and regional streams of the CPE literature to study regional growth regimes. We borrow Hassel and Palier’s (2021) framework and conceptually adapt it to study Italy’s regional economic systems.

Analytical framework: growth regimes meet regionalised capitalism

The construction of our analytical framework proceeds in two steps. We first introduce the main features of the growth regimes approach and argue for a regional approach to study models of capitalism. Secondly, we conceptualise and operationalise the export-led and AK regimes to be analysed in the empirical section.

Growth regimes gone subnational

Baccaro and Pontusson (2016) have popularised the study of growth models within the field of CPE. The approach differentiates post-Fordist models of capitalism based on the main components of aggregate demand: consumption, investment, government spending and net exports. Economic models vary depending on the main contributor to the formation of aggregate demand. A consumption-led growth model is one driven by households’ consumption, upheld for example by real wage growth or by credit-backed consumption (Crouch 2009). An export-led growth model thrives on net exports via current account surpluses. A state-led growth model rests on government consumption through budget deficits. Among these, consumption-led and export-led growth appear to be the two most common models across Europe (Baccaro and Hadziabdic 2023).

Drawing on these insights, Hassel and Palier (2021) made use of the concept of growth regimes to study both the supply- and demand-side of economic systems. Growth regimes are defined as the overarching mode of economic governance geared toward the creation of growth and employment in the national economy and consist of three main components: the engine of growth, the main components of aggregate demand and the institutions organising the economy. The engine of growth refers to the main economic sectors which most contribute to growth and job creation in the economy. The components of aggregate demand – as in Baccaro and Pontusson (2016) – refer to the main demand drivers of growth. The socio-economic institutions – as in VoC – refer to the regimes’ key institutional domains shaping patterns of growth and economic actors’ strategic interactions, namely: the modes of financing firms’ economic activity; product market regulation; the wage-setting system; the skill-formation system and the social protection system.

This theorisation can be applied fruitfully at the subnational level. Studying regional growth regimes, we posit, becomes interesting and necessary under two interrelated conditions, namely:

(1): when marked internal diversity in economic outcomes or productive structures exists across regions within (generally larger) countries;

(2): when subnational governments have key autonomous powers to develop own complementary/alternative socio-economic institutions/policies in support of regional growth regimes.

Under these conditions, methodological nationalism in the study of growth regimes risks yielding confounding results, as epitomised by the Italian case (Locke 1996, Trigilia and Burroni 2009). This is because, under conditions of marked internal diversity in the level of economic development or in regional productive structures, national socioeconomic institutions and public policies which are similar in form will differ in the functions² they execute or the effects they exert locally. Consider the example of a national welfare programme such as Italy’s Citizens’ Income (Reddito di Cittadinanza) – a national social assistance policy launched in 2019 as a form of conditional guaranteed minimum income. While in its form the programme was applied equally across the national territory, its uptake was markedly asymmetric in Italy: by autumn 2020, 61 per cent of the social assistance scheme’s beneficiaries resided in underdeveloped southern Italy, 15 per cent in central Italy and 24 per cent in the north (INPS 2020, p. 4). In other words, a national institution has come to play different socioeconomic functions across heterogeneous regional economies. If in the north the scheme was a vehicle for social inclusion for people at the fringe of the economy, in the south it became a major channel for the state to support household incomes and local consumption at large.

Differently, but relatedly, when subnational governments are endowed with – or obtain – the powers to develop own institutions/policies, those with greater fiscal and administrative capacity are likely to develop policies and institutions that deviate in melius from the rest of the country. This enhances countries’ internal diversity and socioeconomic inequality considering that the economic returns of decentralisation empirically accrue in those territories where local governments’ quality was ex ante already better (Rodríguez-Pose and Muštra 2022). In terms of social and labour market policy and institutions, which are key within growth regimes, subnational governments have evolved from their previous role of residual policy providers toward becoming full-fledged policy actors within multilevel governance systems (Kazepov 2010, Scalise and Hemerijck 2022).

In sum, focusing on subnational growth regimes should be of even greater relevance today considering the three-decade long decentralisation of powers to the subnational level across most European states (Hooghe et al. 2010).

Theorising growth regimes: export-led growth and administrative keynesianism in the Italian regions

To perform our analysis of the regional growth regimes, we draw on Hassel and Palier’s (2021) framework and analyse: (1) the growth drivers of regional demand; (2) the main sectors operating as the regional engine of growth; (3) the institutional setting within which the regional economy is embedded (see Table 1). We focus specifically on three supply-side domains: the welfare domain as the set of (sub)national social policy and skill-formation provisions; the labour market and wage-setting domain; and the corporate finance domain as the prevalent mode of firms’ financing in different territories.

Table 1. Demand and supply side characteristics of export-led and administrative Keynesianism growth regimes

Characteristics of growth regimes	Italy’s two growth regimes
	Export-led	Administrative Keynesianism
Growth drivers	Net exports (X–-M)	Household Consumption (C) + Government spending (G)
Engine of growth	Manufacturing	Public sector + irregular employment (common in low-end services)
Welfare and social policy domain	Regional social investment policies in melius vis-à-vis national provision (VET, ECE, PES)	National consumption-enhancing social policies
Labour market and wage-setting domain	Regional and firm-level bargaining + corporate-based welfare agreements	State as employer of last resort with centralised PSWS + Economic forbearance of labour market regulations (de facto tolerance of irregular work)
Corporate finance domain	Local bank-based credit	Economic forbearance of tax regulations (de facto tolerance of corporate income tax evasion)

Source: Own elaboration.

Italy provides an interesting case study in terms of both conditions posited above. Regarding condition 1, in terms of economic outcomes, Italy displays some of the largest within-country regional disparities across Europe (Balakrishnan et al. 2022) – for example, in the level of per-capita GDP (Figure 1, left panel) and regional employment rates (Figure 1, right panel). Moreover, Italian regions’ productive structures differ substantially, as do the regions’ welfare systems, labour market institutions (Colombo and Regini 2016, Ascoli et al. 2018) and practices of corporate finance (Albareto et al. 2022).

Figure 1. Regional dispersion of GDP per inhabitant in current prices (left panel) and employment rates 20–64 years (right panel) in selected European countries, year 2019. Source: Own elaboration based on Eurostat data.

Regarding condition 2, under the principle of vertical subsidiarity, Italy’s regional governments have acquired greater policy competences through a process of administrative decentralisation culminating in the 2001 Constitutional Reform. The reform transferred important policy competences to regional governments in various subdomains of labour market and social policy, such as vocational education and training (VET) and public employment services (PES). Other social investment policy domains, such as early childhood education (ECE), were instead already decentralised before the reform. Due to weak national coordination, decentralisations have greatly exacerbated regional differences, with northern regions strengthening the territorial provision of public and collective goods as well as their social and economic policies, resulting in greater diversity across regional social and economic models (Colombo and Regini 2016, Ascoli et al. 2018).

Analytically, we identify two growth regimes to be distinguished across Italy’s regions (see Table 1): a manufacturing-based export-led growth regime and one that we theorise as AK. Drawing on Hassel and Palier (2021), the former is a growth regime whose main growth driver lies in net exports originating from a large and competitive manufacturing sector which operates as the key engine of growth and employment in the regional economy. Accordingly, these regions display relatively high trade openness and higher productivity levels among export-oriented industrial producers.

The supply-side institutions and policies underpinning this growth regime are geared toward providing the public and collective goods necessary to sustain territorial competitiveness and industrial firms’ export-oriented strategies. Concerning territorial welfare regimes and skill formation, VET, PES and ECE fall within the remits of subnational governments in Italy and can therefore be singled out in the study of regional supply-side institutions. VET systems are key supply-side institutions shaping firms’ productive strategies (Hall and Soskice 2001) and ensuring the necessary provision of a technical, skilled workforce in support of the manufacturing sector (Benassi et al. 2022). VET in Italy is characterised by a multilevel governance system. The national ministries of education and labour lay out the framework rules. Regions liaise with local social partners to shape active labour market policies and enjoy exclusive legislative powers in governing the planning, organisation and implementation of VET programmes and most apprenticeship schemes (INAPP 2022). Similarly, welfare institutions shape workers’ skill profiles and support firms’ productive strategies (Wren 2020), while social investment policies – such as PES or ECE services – support competitiveness by activating the labour force and guaranteeing its employability through life cycles (Hemerijck 2017). In Italy, the decentralisations of the 1990s assigned to regional governments the competence for the governance of PES and active labour market policies.³ ECE services for children 0–2 in Italy have long been the responsibility of regions which refer to their municipalities for planning and implementation. This has historically resulted in marked territorial differences between the northern territories where coverage is high and southern Italy where the provision of ECE services has lagged (Da Roit et al. 2019).

Regional labour market policies and wage-setting institutions can contribute to enhance the competitiveness and attractiveness of regional firms (Trigilia and Burroni 2009). In 1993, the Italian collective bargaining system was reformed to introduce a two-tiered wage-setting system. The new system combines national-level sectoral bargaining (setting floor wage increases based on forecast inflation) with territorial and firm-level agreements topping the national tier based on territorial or firm-specific productivity gains (Di Carlo 2023). Decentralised bargaining allows companies to exchange local wage moderation and/or firms' internal flexibility for supplementary company-based or territorial welfare provisions. These local welfare provisions are provided to workers and their families in liaison with other regional stakeholders (e.g. non-profit organisations). Subsidised by national tax incentives, decentralised bargaining has grown steadily over the years and is now more common in northern industrial regions where competitiveness-enhancing corporate agreements are often integrated within territorial social pacts including the involvement of local administrations (Ferrera and Maino 2014, Ascoli et al. 2018).

Not less important, firms need access to capital for their operations, and the characteristics of corporate governance and financing systems shape both firms’ productive strategies and innovation patterns (Hall and Soskice 2001, Simoni 2020). Italy’s ‘relationship-based’ corporate governance system is characterised by the limited role of the stock market and, as a result, self-financing and bank-based credit as firms’ main source for corporate funding (Melis 2000, p. 350). However, differently from Germany, companies’ management is not the banks’ business, provided firms refund their debts. Rather than disappearing after the 1990s privatisation of the Italian banking system, firms and banks have in fact become more tightly intertwined at the regional and local level in Italy (Jones 2021, p. 436). But only within the industrial districts have banks generally developed close ties with SMEs and engaged in long-term relations typical of the German bank-based, relational system (Deeg 2009, p. 11).

Our theorised AK regime shares major traits with Hassel and Palier’s (2021) ‘publicly-financed domestic demand-led growth regime’, where domestic consumption is upheld by generous public spending. Countries in this regime, to be found across southern Europe, tend to have large, sheltered sectors (Scharpf 2016) and high levels of ‘consumption-enhancing’ social policies (see also Beramendi et al. 2015) aimed at supporting household consumption. This theorisation, however, focuses only on social provision and overlooks the other channels through which the state can (in)directly support households and firms through its regulatory/enforcement powers or through its political role as employer of last resort in the public sector.

Thus, we theorise AK as a growth regime where the main growth driver lies in domestic consumption upheld by compensatory social policies and (in)direct forms of transfers to regional residents. Government spending is the foundation of such a regime and the predominant source backing households’ capacity to consume. For lack of developed alternatives, two sectors predominantly make up the regional engine of growth in the AK regime, namely the public sector and low-end private services, where the likelihood of finding large segments of irregular employment is high. These economies remain relatively sheltered from international trade and display much lower productivity levels, which in turn also contributes to slower economic growth and lower own fiscal revenues for southern regional governments.

The supply-side institutions and policies of the AK regime are geared toward channelling public resources to households and firms to support local demand and employment. Here, the concept of AK aptly captures the underlying logic of the regime. Crouch has characterised privatised Keynesianism as an unacknowledged policy regime where ‘instead of governments taking on debt to stimulate the economy, individuals did so,’ and did so thanks to the ‘growth of credit markets for poor and middle-income people, and of derivatives and futures markets among the very wealthy’ (Crouch 2009, p. 390). In the AK regime, it is the state’s budgets and public debt – rather than private financial institutions and household debt – which stimulate demand among the poor and middle-income people. However, AK is not simply a Keynesian counter-cyclical fiscal policy. Along with Crouch, we conceptualise AK as a policy regime where the state operates functions of demand management through various channels in the welfare, the labour market and the corporate finance domains.

First, in the welfare domain, AK regions’ weak fiscal and administrative capacity prevents them from developing territorial competitiveness-enhancing institutions and policies. These shortcomings are only partly compensated for by national social policies which, by virtue of the significant territorial differences, have a higher uptake and exert stronger compensatory effects in Italy’s underdeveloped regions. In Italy, this is the case most prominently for two major items of national social spending, i.e. disability or survival pensions, which include pensions not paid out of defined contributions, and the Citizens’ Income. These schemes’ uptake is highly asymmetric across the country, with a more concentrated number of beneficiaries in the AK regions where it contributes to bolster the consumption capacity of a larger number of local beneficiaries.

Second, in terms of labour market policies, for lack of attractive employment opportunities in the manufacturing sector and high-end services, the state can pursue two types of employment policies. On the one hand, it can serve as the employer of last resort, directing jobless residents into public employment. In underdeveloped areas, a relatively larger share of public employees does not necessarily result from high numbers of per-resident public employees. But it can simply result from low overall employment, given the underdevelopment of industry and high-end services, and a large NOE. The expansion of the state’s function of employer of last resort in Italy has occurred at least since the institutionalisation of regional governments in the 1970s (Santoro 2014). This role has been reinforced by the system of centralised public sector wage setting (PSWS) (Di Carlo 2023), which provides uniform nominal wages and wage increases across territories with marked north/south differences in economic development and the cost of living. In Italy, both public employment and centralised public sector wage setting have been used as a hefty subsidy to redistribute fiscal resources in support of the less wealthy southern regions (Alesina et al. 2001), thereby upholding southern households’ incomes.

On the other hand, the state can informally support employment by ‘turning a blind eye’ to the enforcement of labour market regulations. This practice is generally understood as economic forbearance, through which states can selectively aid groups of economic producers by de facto tolerating the non-enforcement of costly economic regulations (see Dewey and Di Carlo 2022). Thus, states can indirectly subsidise employment creation in underdeveloped areas by tolerating irregular employment. But economic forbearance has fiscal costs and it is a form of AK because the state supports employment in the NOE not through direct employment subsidies but through state inaction, i.e. by de facto tolerating the foregone fiscal revenues from the labour taxes and social contributions employers and employees would have paid, had employment been regular.

Third, in terms of corporate finance, while firms in northern export-led regions can leverage their ties with local banks, southern Italy has a less developed banking system and southern firms’ access to bank-based credit is both more difficult and more costly (Albareto et al. 2022). The weaknesses of the local banking system incentivise corporate tax evasion through which firms can turn tax payments withheld from the state into corporate self-financing or simply profit-making. For the state, tolerating the non-enforcement of corporate tax regulations becomes a functional equivalent to corporate subsidies, i.e. an indirect subsidy to local producers in underdeveloped areas to keep them afloat amidst the broader deficiencies of the economic system (Dewey and Di Carlo 2022). Accordingly, tolerating irregular employment and corporate tax evasion can be understood as two complementary sides of economic forbearance through which states can support the local economy within the AK regime.

Between export-led growth and administrative Keynesianism: Italy’s regional growth regimes

Guided by the analytical framework introduced above, our empirical analysis unfolds in three steps. We present data on: (1) the regional growth drivers and patterns of economic growth; (2) the main regional economic sectors (engines of growth); and (3) the characteristics of the supply-side institutions. Our aim is analytical/descriptive and our purpose is to identify clusters of Italian regions belonging to – and displaying the characteristics of – the two growth regimes which characterise Italy’s two-tiered political economy. Readers are referred to the appendix for more detailed information about data sources and methodology.

Regional growth drivers and economic growth in Italy

Two clear-cut growth regimes can be identified in Italy’s national political economy (Figure 2). The northern regions belong to the export-led growth regime, i.e. Emilia-Romagna, Friuli-Venezia Giulia, Lombardia, Piemonte, Toscana and Veneto (plotted in black in the figures). Vice versa, the southern regions can be ascribed to the AK regime, i.e. Calabria, Campania, Puglia, Sardegna and Sicilia (in lighter grey).

Figure 2. Export share and final consumption (public and private) as percentage of regional GDP in Italian regions, average values 2015–2019. Source: Own elaboration based on ISTAT data, territorial accounts, and ISTAT-Ice yearbook. Note: Axes represent national average values for the selected variable. Exports represent the region’s exports to the rest of the world.

The northern export-led regions display a very high share of exports over regional GDP, at around 35–45 per cent (Figure 2). These are the regions where the contribution of net exports to GDP growth is relatively higher than the rest of the country during the period observed (Figure 3). However, it must be noted that the growth contribution of net exports is modest (or negative in same cases), reflecting the overall weaknesses of Italy’s stagnating national growth model (Baccaro and Pontusson 2016). The southern regions are characterised by very low export shares, high levels of regional (private and public) consumption (Figure 2), and a negative growth contribution of net exports to regional GDP (Figure 3).

Figure 3. Contribution to GDP growth of net exports and ratio between export and consumption share in the regional economy, Italian regions, average values 2015–2019. Source: Own elaboration based on ISTAT data, territorial accounts, and ISTAT-Ice yearbook. Note: Axes represent national average values for the selected variable. Exports represent the region’s exports to the rest of the world.

Figure 4 points to diachronic and territorial variation in patterns of economic growth. Overall, during the 2000s the southern AK regions grew at a much slower pace than Italy’s average and the northern export-led regions.⁴ During the aftermath of the financial crisis (2008–2014), overall economic growth in Italy was negative, but degrowth was much more pronounced in the AK regions than in the export-led ones. In the years before the Covid-19 pandemic (2015–2019), growth picked up timidly across the country but remained very low in the southern AK regions.

Figure 4. Real GDP average growth rate by region. Chain-linked volumes (ref. 2015). %. Years 1999–2007, 2008–2014, 2015–2019. Source: Own elaboration based on ISTAT data, territorial accounts.

In sum, most of the Italian regions cluster by and large around the two growth regimes. Northern export-led regions tend to have higher growth rates, while the southern AK regions based on regional consumption display much lower growth rates. The few remaining regions (Abruzzo, Lazio, Liguria, Marche, Umbria) evade straightforward categorisation and are therefore considered hybrid and excluded by the following analysis (for other minor regions excluded, see the appendix).

Regional engines of growth

The northern export-led regions have large manufacturing sectors (20–25 per cent of regional VA) and have a relatively modest share of public sector VA (Figure 5). Moreover, the share of the NOE over regional GDP is substantially smaller in the northern export-led regions, as indicated by the size of the bubbles in Figure 5. The southern AK regions have much larger public sectors (25–30 per cent of regional VA) and a much greater incidence of the NOE in the regional economy.

Figure 5. Sectoral gross value added and the shadow economy in Italy’s regions, average values 2015–2019. Source: Own elaboration based on ISTAT data, territorial accounts. Note: Axes represent national average values for the selected variable.

Such differences are evident when considering regional productive structures and specialisations. Northern Italy is characterised by a manufacturing-based economy, with a territorial specialisation in the metalworking, engineering and chemical sectors with a larger share of multinational groups and where larger firms operate alongside SMEs as the major economic players in a resilient manufacturing sector (SVIMEZ 2023, p. 129). The northern regions of the export-led variety are thus akin to small open economies, with higher levels of international trade openness (Figure 6, left panel) and higher labour productivity (Figure 6, right panel).

Figure 6. Regional trade openness index (imports + exports) as percentage of regional GDP (left panel) and productivity (right panel), selected Italian regions, averages 2015–2019. Source: Own elaboration based on ISTAT-Ice yearbook and ISTAT data, territorial accounts. Note: White regions represent minor, or ‘hybrid’, regions excluded from the analysis. Productivity is measured as GVA per hour worked.

On the contrary, southern AK regions have remained relatively sheltered economies (Calabria’s trade openness is just 5 per cent) with sectoral specialisations in the public sector and low-end private services and with small and micro-firms operating in sectors such as wholesale/retail trade and hospitality (SVIMEZ 2023, p. 129), prone to fall prey to economic informality in the NOE.

The supply-side institutions

Turning to the supply side, the empirical analysis focuses on three domains (see Table 1): the welfare, the labour market and wage-setting and the corporate finance domains.

To analyse the welfare domain across regions, our point of observation is twofold: we gauge the extent to which ensembles of regional social policies/institutions upholding manufacturing-based territorial specialisations are diffused across Italian regions. We also analyse the territorial distribution of beneficiaries of those national consumption-enhancing social policies instrumental to supporting regional demand in the AK model.

Based on the analytical dimensions introduced in Table 1, we created a synthetic index of social investment services comprising those social investment and skill-formation institutions under the remit of Italy’s regional governments (VET, PES, ECE). To analyse the territorial diffusion of consumption-enhancing social policies, we focus on the number of regional beneficiaries of the two major national social policy provisions, i.e. the Citizens’ Income and welfare pensions.

Figure 7 makes explicit the territorial divide which marks the supply-side of the two opposite growth regimes. Northern regions benefit from the presence and greater diffusion of competitiveness-enhancing institutions in support of the manufacturing export-led growth regime. These regional welfare provisions are absent – or at best underdeveloped – across the southern AK regions. On the contrary, in AK regions there is a much larger share of residents benefitting from national social compensation schemes, funded by national budgets and de facto operating to support local households’ capacity to consume. It is important to highlight, however, that the lack of developed regional institutions in AK regions did not necessarily come about by choice. It is often dictated by these regional governments’ incapacity to raise own fiscal resources to fund complementary social policy provisions. Table A3 in the appendix shows how the AK regions generate own per-capita fiscal resources that are much lower than the national average and the export-led regions. With weaker regional economies – and a larger NOE – AK regions are stuck in a bad equilibrium where weaker fiscal capacity is only partially compensated for by the national taxation system. As a result, AK regions can afford lower per-capita expenditures than the rest of the country. This results in a vicious circle and increases the dependence of southern regions on the central government’s redistributive channels.

Figure 7. Synthetic indexes of social investment policies and administrative Keynesianism policies in Italian regions, various years. Source: Own elaboration based on ISTAT, INPS, Inapp, Indire, Anpal data. Note: See Table A1 and A2 in Appendix for further details on the construction of the indexes.

Turning to the labour market and wage-setting, we first focus on the territorial dimension of wage-setting and social partnership in support of manufacturing-based productive strategies. Secondly, we analyse the structure of regional labour markets, focusing on the public sector and the share of irregular employment in regional economies.

Figure 8 neatly indicates the more widespread presence of territorial and corporate welfare agreements in northern export-led regions characterised by the presence of internationally exposed manufacturing firms negotiating territorial social pacts with local economic stakeholders aimed at both territorial competitiveness and workers’ well-being. These provisions, in turn, guarantee further territorial competitive advantages and the attractiveness of regions for both skilled workers and industrial firms. On the contrary, corporate and territorial welfare agreements are hardly observable in southern AK regions.

Figure 8. Distribution of corporate plus territorial agreements across Italian regions, standardised as agreements per 1.000 operating firms in 2020. Source: own elaboration based on ISTAT and Italian Ministry of Labour and Social Policy data. Note: White regions represent minor, or ‘hybrid’, regions not analysed in the paper.

The labour market in southern Italy’s AK regions is structured around the public sector and the widespread presence of irregular employment. While in absolute terms public employees per capita are often lower in southern regions, in relative terms the share of public employees over total employment is much higher (Figure 9) due to very low employment levels and a large share of irregular employment in the regional NOE. A larger public sector in southern Italy reflects the role long played by the Italian state as employer of last resort, thanks to which individuals from underdeveloped areas have been employed by the state apparatus for lack of better employment alternatives in the private sector (Santoro 2014). While being relatively low in northern regions, irregular employment in southern AK regions reaches extremely high levels in Calabria, Campania, Sardegna and Sicilia (Figure 9). The spread of irregular employment is tightly linked to the productive structure of AK economies, centred around low-end private services (and agriculture) along with the public sector. According to ISTAT (2020, p. 3), as of 2018 the incidence of irregular employment in Italy was negligible in the manufacturing industry (between 1.2 per cent and 3 per cent of sectoral VA), while it was the highest in the occupations linked to services to people (22.5 per cent of sectoral GVA) and in agriculture (17 per cent), the sectors that are chiefly overrepresented in southern AK regions.

Figure 9. Share of irregular and public employment as a % of total employment in Italian regions, 2017 and average levels 2015–2019. Source: Own elaboration based on ISTAT data, territorial accounts and second Permanent Census of Public Institutions. Note: Axes represent national average values for the selected variable.

In terms of corporate finance, we proxy the diffusion and operational capacity of Italy’s bank-based system across regions through the presence of bank branches (Figure 10, right panel) and their lending activities to non-financial corporations (Figure 10, left panel). In 2018, 40 per cent of the country’s total bank branches were concentrated in Emilia-Romagna and Veneto. Southern Italy had only 22 per cent of the country’s total branches (Bank of Italy 2019, p. 2). Indeed, Figure 10 (right panel) displays the greatest density of bank branches being in northern Italy, while also indicating that bank-based credit to non-financial corporations (left panel) is substantially lower in southern AK regions.

Figure 10. Local banks’ lending to non-financial corporations as % of regional GDP (left panel) and bank branches for 100.000 inhabitants (right panel), selected Italian regions 2015–2019 and 2019. Source: Own elaboration based on ISTAT data, Indicators for development policies database, Bank of Italy, Banks and financial institutions, territorial structure. Note: White regions represent minor, or ‘hybrid’, regions not analysed in the paper.

Concomitantly, southern Italy’s business undertakings resort to much greater levels of corporate tax evasion. Official estimates by the Italian Ministry of Economy and Finance highlight great territorial variation in the propensity of firms to evade Italy’s corporate income tax (IRES). The incidence of corporate tax evasion over the total firms’ tax base during the period 2014–2019 ranges from 2.5 per cent among some northern regions to almost ¼ of firms’ tax base in southern regions such as Calabria and Sicilia (MEF 2023, p. 26). This suggests that, also due to lack of access to bank-based credit, many firms in southern Italy resort to both corporate tax evasion and irregular employment to hold greater financial resources for corporate self-financing and profit-making.

Conclusions

This article has argued that CPE approaches to the study of growth models/regimes can be enriched by taking into consideration the spatial dimension of growth. Analysing subnational growth regimes becomes relevant when countries are characterised by marked internal diversity in economic outcomes/productive structures or when subnational governments dispose of powers to develop own territorial institutions and policies which diverge from the national level. Under these conditions, the regional growth regimes approach developed here can contribute to a better understanding of a country’s political economy, as epitomised by the Italian case.

The CPE literature focusing on national models of capitalism struggles to categorise Italy’s model of capitalism between a hybrid mixed-market economy and a dysfunctional case. Our analysis has identified Italy as a two-tiered growth regime with two diametrically opposed regional growth regimes which coexist (uneasily) in the national political economy. Italy’s northern regions are manufacturing-based, export-led economies, integrated into international trade, relatively productive and vibrant, and underpinned by functional regional institutions which complement local productive strategies and enhance territorial competitiveness. This explains Italy’s resilient manufacturing prowess and the country’s recurring – yet modest – trade surpluses despite the adoption of the European single currency. On the contrary, southern regions conform to what we have theorised as an administrative Keynesianism (AK) regime with a large informal economy, relatively larger public and irregular employment, one that is less productive and more dependent for growth and employment generation on the state’s role as employer of last resort, its consumption-enhancing social policy programmes, and the forbearance of labour and corporate tax regulations.

The regional approach developed in this article carries some interesting questions and implications for CPE debates. The first pertains to the implications of growth models’ regional heterogeneity for the politics of growth. To the extent that marked structural differences persist across regional growth regimes, regional economic stakeholders will maintain different material interests and regime-specific economic policy preferences. Future research could investigate how heterogeneous regional growth regimes translate into territorial growth coalitions and how the latter shape the politics of (non)growth at the national level.

The second relates to the differentiated impact of European economic and monetary integration on countries with diverse regional growth regimes. Italy’s experience suggests that northern Italy’s export-led regions have managed to cope with greater global competition and a ‘hard-currency’ monetary regime in Europe. Indeed, as of 2022, Italy accounted⁵ for 19 per cent of total EU industrial production, second only to Germany (at 26 per cent). Thus, thirty years after Richard Locke’s writings (Locke 1995), Italy is still characterised by a dysfunctional national political economy hiding resilient patterns of regional economic dynamism. On the contrary, the sheltered AK regions of southern Italy have suffered disproportionately from the Eurozone’s fiscal constraints and the austerity measures implemented after Italy’s sovereign debt crisis. Given the key compensatory role played by public employment in AK regions, the public sector wage freezes and cuts implemented during 2009–2016 (Di Carlo 2023) have eroded southern households’ capacity to consume, further depressing regional demand and growth in AK regions. Likewise, the collapse of public investment observable in Italy over the last decade has penalised southern regions disproportionally, considering these regions’ much lower endowment of public infrastructures (SVIMEZ 2023, p. XVI). Thus, taking regional growth regimes into account points to a more differentiated impact of European integration across heterogeneous territories within national models of capitalism.

Thirdly, unearthing and understanding the structural differences across regions’ growth regimes shall be instrumental for the design of place-based policies to avoid designing ‘one-size-fits-some’ forms of national industrial policies. Related to these aspects, through regional input-output data, future CPE research could also investigate the interlinkages between regional growth models to analyse intra-country patterns of regional trade and value chains and detect complementarities between regional growth regimes.

Considering that the topic addressed in this article is politically contentious in Italy, we conclude by stressing that our analysis carries no normative judgement against southern Italy’s ‘subsidised’ growth regime. Our aim was simply to analyse the ideal-typical characteristics of Italy’s regional growth regimes. Moreover, it should be noted that important manufacturing clusters – e.g. in automotive, clean technology and semiconductor sectors – exist also in southern Italy. They are certainly relevant for local growth and employment and are also home to some excellent firms. However, they remain too small and fragmented to significantly alter the major traits of these southern AK regions.

Acknowledgements

A previous version of this paper was presented at the Max Planck Institute for the Study of Societies in Spring 2023. We are thankful to all participants for their comments and helpful insights. For their instructive suggestions, we are indebted to Lucio Baccaro, Sinisa Hadziabdic, Martin Höpner, Marco Oberti, Marco Simoni, Arianna Tassinari, and Dustin Voss.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Donato Di Carlo

Donato Di Carlo is a lecturer in political economy at Luiss University in Rome, Italy, and managing director of the Luiss Hub for New Industrial Policy (LUHNIP). His research focuses on comparative political economy, industrial policy, and industrial relations.

Andrea Ciarini

Andrea Ciarini is Associate Professor of Economic Sociology at the Department of Social and Economic Sciences of the Sapienza University of Rome where he specialises in Economic Sociology and the Sociology of Welfare.

Anna Villa

Anna Villa is a researcher at the President's Office of the Italian National Institute of Statistics and has recently obtained her PhD in Applied Social Research at Sapienza University of Rome. Her research focuses on well-being indicators and welfare policies.

Notes

1 According to ISTAT, the NOE includes ‘underground activities, illegal activities and other productive activities which are missed due to statistical reasons’.

2 On the difference between institutions’ form and functions, see Streeck and Thelen (2005).

3 As part of the so-called ‘Bassanini Reforms’, from the name of the Minister for Public Function and Regional Affairs who carried out Italy’s administrative decentralisation.

4 However, consider that even Italy’s most developed territories, such as in the northeast, lag very much behind other major European economies, see also Krahé (2023, p. 10).

5 Based on Eurostat, Industrial Production Statistics.

Appendix: data and methodological remarks

The choice to analyse regional growth regimes for the period 2015–2019 is motivated by two considerations. First, to study growth patterns, we had to select a period of economic growth. Since GDP growth was, on average, negative during the period 2008–2014, our choice would have been for either the years preceding the great financial crisis (1999–2008) or the period 2015–2019. However, regional data is scant for the former period, leaving us no choice but to focus our analysis on the latter period. We have excluded from the sample some of Italy’s minor regions, either for lack of data (namely Trentino-Alto Adige, Südtirol and Valle d’Aosta) or because of their negligible weight in the national economy (Molise and Basilicata). When excluding these territories, in 2019 the set of regions included in the sample represents about 96 per cent of Italy’s overall GDP. The 'hybrid regions' identified in the paper (Abruzzo, Lazio, Liguria, Marche, Umbria) are included in the data but excluded from the analysis in the empirical section. Admittedly, this is a noteworthy limitation of our paper and future research could investigate in greater details the characteristics of these hybrid regions' economic systems.

Unless specified, the empirical analysis is based on official data sources compiled by the Italian Statistical Office (ISTAT). All data concerning the demand drivers and the engines of growth refer to 2015–2019 and are average values at current prices (except for the contribution of net exports to the GDP growth, which is computed using both current and previous year prices consistently with national accounts). For the institutional setting, data refers to the pre-pandemic year (2019) except for pensions which are 2015–2019 averages and of public employment which refers to 2017 (yearly time series are not available). The empirical analysis develops through three steps.

First, we analyse the growth drivers in Italy’s regions, identifying the main components of aggregate demand at the regional level and thereby assigning regions to their respective growth regime, be it export-led or administrative Keynesianism (AK), as elaborated in the theoretical section of the paper (see also Table 1 in the paper). In Figure 2, we analyse the share of the main demand components over the regional GDP, namely the export share (including both goods and services to the rest of the world) and the final consumption expenditure share, which includes both public (general government) and private sector consumption (households and non-profit institutions serving households). In Figure 3, we plot the contribution of regional net exports to regional GDP growth (horizontal axis) against the ratio between the share of regional exports over GDP and the share of regional consumption (public and private) over regional GDP (vertical axis). We assign regions to the two clusters based on two criteria, one structural, i.e. based on the relative size of the demand side components over regional GDP and one centred on flows, i.e. on the calculation of the average contribution to GDP growth of the specific demand component. We proxy the size of the demand components by computing the ratio between export and consumption (public and private), attentively arriving at a measure of the relative importance of the export sector.

Through regional accounts, we then calculated the average yearly contributions to GDP growth of net exports, using the following equation: (1) $\mathrm{Contra}(t)={\displaystyle \frac{[\mathrm{prean}(t)-\mathrm{coran}(t-1)]}{\mathrm{PILcoran}(t-1)}}$(1) where Contra(t) = contribution to GDP growth of a generic annual series in the year t; prean(t) = generic annual series in the year t in previous year prices; coran(t − 1) = generic annual series in the year (t − 1) in current prices; GDPcoran(t − 1) = annual GDP in the year (t − 1) in current prices.

Accordingly, the hybrid regions, mostly located in the centre of Italy that evade straightforward categorisation, have been excluded from the subsequent fine-grained analysis, i.e. Abruzzo, Lazio, Liguria, Marche, Umbria. After having assigned regions to the two growth regimes, in Figure 4 we plot the growth patterns for the various growth regimes over the period 1999–2019, displaying real GDP growth.

Second, after having assigned the regions to their specific growth regimes, we present granular and original data on regions’ main economic sectors which constitute the engines of growth. In Figure 5, we include the manufacturing sector’s value added (VA) (excluding construction) and the VA of the public sector. Composed of the following subsectors: public administration and defence, compulsory social insurance, education, health, and social assistance. We furthermore include estimations of the VA deriving from the non-observed economy (NOE) as a percentage of total regional GVA, which, according to ISTAT, includes ‘underground activities, illegal activities and other productive activities which are missed due to statistical reasons.’ To further uncover the structural differences across subnational growth regimes, in Figure 6 we also analyse regional trade openness (calculated as the sum of imports and exports over regional GDP) and productivity levels (computed as GVA per hour worked). The former indicator provides a standard proxy for the economic openness of regions and their integration in international markets. The latter represents a key measure of regional competitiveness.

Third, we analyse the supply side institutions which structure regional economies. We focus on the nexus between social protection and labour market policies, distinguishing between social investment policies and ‘consumption-enhancing’ policies, as defined in the article’s theoretical section. To capture the presence and diffusion of each type of policy/institutional ensemble, we built two distinct indexes, one meant to proxy social investment policies and the another to gauge consumption-enhancing social policies.

To gauge the territorial diffusion of social investment services at the regional level, we have built a synthetic index (shown in Figure 7) starting from 4 basic indices (see Table A1 for details): provision of childcare services (ECE), the supply of vocational educational training (VET) courses (upper secondary and post-secondary level) and the availability of public employment services (PES). In short, all these indicators represent policies relevant from a territorial point of view because they are a regional or sub-regional competence (as in the case of early childhood education services) and cover most of the salient domains in terms of social investment.

To obtain the synthetic index, data have been standardised through the formula: (2) $z={\displaystyle \frac{(X-\mu )}{\sigma }}$(2)

The final synthetic index has been calculated as an arithmetic mean of the standardised values: it has zero average and assumes positive values where the presence of services on the territory is more developed, and negative where they are missing.

The other measure used in Figure 7 to gauge the diffusion of ‘consumption-enhancing’ social policies across regions is the analysis of two social programmes attributable to a compensatory welfare model aimed at supporting consumption: disability or survival pensions, which include all pensions that are not paid out of defined contributions, and the Citizens’ Income (Reddito di Cittadinanza). In this case, the index was calculated as the sum of the recipients in relation to the resident population and provides a measure of how many people in the different regional territories benefit from these programmes (see Table A2).

Figure 8 maps the regional distribution of firm-level and territorial agreements, complementing national collective bargaining. This measure is standardised as the number of agreements per 1000 operating firms (in 2020). Data on firms comes from Istat, while data on subnational agreements comes from the Ministry of Labour and Social Policy.

In Figure 9, on the horizontal axis, we plot the share of public employment as the percent of regional total employment, useful as a proxy of the role of the state as an employer of last resort within the different regional labour markets. On the vertical axis we plot the share of irregular employment in total regional employment, yearly provided by ISTAT in the territorial account. Estimates are available at https://noi-italia.istat.it/ (in Italian).

Lastly, Figure 9 aims to gauge the regional distribution of bank-based credit and the presence of local banks. The left panel plots local banks’ lending to resident non-financial corporations as per cent of regional GDP (during 2015–2019, average), while the right panel plots the number of bank branches for 100,000 inhabitants across the selected Italian regions in 2019. The former indicator comes from the Indicatori territoriali per le politiche di sviluppo Database (in Italian), the latter is yearly published by Banca d’Italia (the Italian central bank) in the report ‘Banche e istituzioni finanziarie, articolazione territoriale’ (2020 edition).

Table A3 provides the per-capita own fiscal revenues and expenditures of Italian regions in 2019 (at current values), based on data from the Italian Agency for Territorial Cohesion (Agenzia per la Coesione Territoriale – Conti pubblici territoriali (CPT), December 2022 edition).

Table A1. Indicators used for the Social Investment Services Index.

SI dimension	Indicator	Unit	Source	Period
Childcare	Places in the publicly provided ECE services	For 100 children aged 0–2	Istat	2019
VET courses	Upper secondary-level courses	For 1000 young people up to 24 with lower secondary education	Inapp and Istat (own elab)	a.f. 2019/2020
	Post-secondary level courses	For 10,000 young people up to 24 with upper secondary education	Indire and Istat (own elab)	May 2020
Public employment services (PES)	Availability of PES	For 10,000 unemployed aged 20–64	Anpal and Istat (own elab)	2019

Table A2. Indicators used for the Consumption-enhancing Social Policies Index.

Indicator	Unit	Source	Period
Beneficiaries of welfare pensions	Per 100 inhabitants	Istat (own elab)	2015–2019
Inclusiveness of the RdC	Per 100 inhabitants	INPS and Istat (own elab)	2019

Table A3. Per-capita own fiscal revenues and expenditures of Italian regions in 2019 (current values).

	Per-capita fiscal revenues from own resources			Per-capita fiscal expenditures
Region	Local administrations	Regional administrations	Total subnational administrations	Local administrations	Regional administrations	Total subnational administrations
Piemonte	823	900	1723	1051	2498	1723
Valle d'Aosta	959	1285	2244	1720	5403	7123
Lombardia	781	949	1730	1122	2096	3218
P.A. Trento	509	898	1406	2016	5309	7325
P.A. Bolzano	680	921	1600	2570	8395	10964
Veneto	748	782	1530	1023	2183	3206
Friuli Venezia Giulia	526	693	1219	1502	3883	5386
Liguria	1097	846	1943	1341	2411	3752
Emilia Romagna	898	885	1783	1124	2326	3450
Toscana	910	811	1721	1183	2069	3252
Umbria	847	557	1404	1121	2331	3451
Marche	767	681	1449	1171	2242	3413
Lazio	885	1052	1937	1166	2154	3320
Abruzzo	720	743	1463	1540	2179	3719
Molise	667	634	1301	1313	2270	3583
Campania	721	534	1255	895	1793	2688
Puglia	694	417	1111	894	2051	2946
Basilicata	626	796	1422	1277	2388	3665
Calabria	623	557	1180	1030	2045	3075
Sicilia	619	529	1148	921	2721	3643
Sardegna	621	422	1043	1463	2655	4118
Italy	771	764	1535	1118	2335	3453

Source: Authors’ elaboration based on data from the Italian Agency for Territorial Cohesion (Agenzia per la Coesione Territoriale – Conti pubblici territoriali (CPT), December 2022 edition).