Round Table. Implementing the EU Directive on adequate minimum wages in southern Europe: the odd case of Italy

Introduction

If one looks only at the data used by the European Commission and the OECD one might come under the false impression that southern European countries do not need to make much effort to comply with the aim (established in Article 4, para 1 of the Directive on adequate minimum wages) of promoting collective bargaining on wages. Collective bargaining coverage is already very high in all of these countries, and particularly so in Italy, where it is estimated by international institutions to be the highest in the EU, at the unbeatable level of 100 per cent (in France it is reported at 98 per cent, in Spain at 80 per cent and in Portugal at 74 per cent; OECD-AIAS, 2022). Nonetheless, these countries also face the problem of in-work poverty (Lohmann and Crettaz, 2018), and their wage regulations are not waterproof, as experienced during the turbulent reforms and unilateral government interventions of the past decade (Meardi, 2014). In the extreme case of Greece, coverage collapsed from around 80 per cent to under 50 per cent following recent internationally mandated structural reforms.

In fact, the ‘southern European’ cluster of industrial relations systems, which is often associated with strong state regulation and politicised industrial relations processes (European Commission, 2009) displays a particular heterogeneity, which goes beyond the often-remarked fact that national clusters hide deep divides, for example, by sector (Bechter et al., 2012). If southern European countries share some institutional and cultural legacies, such as late and uneven industrialisation and, in the Catholic ones, a Church–state conflict that has fostered politically divided trade unions (Crouch, 1993), these countries’ paths have diverged in some very important respects. Therefore, if at first sight collective bargaining is centralised and state-supported in all these countries, its incidence and legal underpinnings are very different: the countries have different ‘regimes’ of collective bargaining extension (García Calavia and Rigby, 2020; Hayter and Visser, 2018); and derogations from the favourability principle have been introduced and repeatedly amended in all of these countries, at different times and with different rules. As a result, the incidence of sectoral collective bargaining on wages and working conditions varies country by country in terms of both the lowest wages and enforceability. In particular, the mechanisms for legal extension are distinct. It takes place automatically by legislation in Spain, by government decisions in France and Portugal, and through the courts with the apparent ‘functional equivalent’ of an established interpretation of the Constitution in Italy. From this variation in form and incidence derives the fact that the actual challenges arising from the EU Directive are also different.

In this group, Italy is particularly distinctive. In the field of labour law, it shares some aspects of the German path after the Second World War, notably the protection of collective bargaining autonomy against government interference, as a radical break from the totalitarian period. In Italy, until the 2010s, a statutory minimum wage was not even considered as an option. Moreover, the principle of legal extension that characterised Fascist-period ‘corporative contracts’ was rejected. Associational regulations on industrial relations in Italy are therefore much more widespread than state regulations, far more so than in other larger EU Member States (Meardi, 2018).

As a result, Italy today is in a unique situation among southern EU countries in not having a national minimum wage, nor a legal framework for collective bargaining extension. In this article we will examine the main problems of the current Italian wage-setting system, concluding that in order to promote adequate minimum wages effectively in accordance with the aim of the Directive, some intervention is needed. The article concludes by outlining two possible lines of reform that Italy can consider in order to comply with the aim of promoting collecting bargaining on wages, as prescribed by Article 4, para 1, for all EU Member States regardless of current coverage, and with the requirement of Article 10, para 2 to report in detail on collective agreement data.

Neither statutory, nor voluntaristic: the dilemma of Italian wage setting

The Italian collective bargaining system is organised autonomously by the social partners and has long been centralised at the sectoral level. Its current form derives from a tripartite agreement (the so-called ‘Ciampi protocol’) of 1993 and has been confirmed with updates by more recent centralised bipartite or tripartite agreements, most lately the bipartite ‘Factory Pact’ of 2018. This autonomous system relies on the centrality of national sectoral collective agreements (Contratti Collettivi Nazionali di Lavoro, thereafter CCNL) that cannot be derogated from by lower-level agreements in relation to wages. After the peak of centralised, egalitarian collective bargaining with the flat-rate wage indexation system agreed in 1975, a process of decentralisation has gradually occurred since the 1980s. But this process has always encountered a barrier in Italy’s fragmented, geographically diversified production system, in which small enterprises are prevalent and traditional managerial practices (especially in family-owned companies) remain at odds with collective bargaining (Regalia, 2019). As a result, as in Spain and Portugal, Italian employer organisations have always resisted the temptation to abandon centralised collective bargaining, even when there have been opportunities, as during the Eurocrisis of 2010–2012 (Bulfone and Afonso, 2020). The most disruptive attempt at decentralisation occurred in 2010–2011, with Fiat-Chrysler’s exit from sectoral collective bargaining and the subsequent decision by the Berlusconi government to introduce the principle of general enforcement of company-level agreements with Article 8 of Decree 138/11. But even in that case, wages were explicitly excluded from the possibility of derogating from national sectoral agreements. In other words, Italy never had something as far reaching as the ‘Pforzheim agreement’ that facilitated derogations in the German metalworking industry.

Centralisation and coverage have no support in Italian statutory law, however. The prominence of sectoral agreements derives instead from the established orientation of the courts, since the 1950s, towards referring to national collective agreements in order to determine what ‘commensurate’ remuneration is when enforcing Article 36 of the Italian Constitution (Menegatti, 2019):

Workers have the right to remuneration commensurate with the quantity and quality of their work and in any case such as to ensure them and their families a free and dignified existence.

Because all Italian workers have the right to claim redress if they are not paid collectively agreed wages, many analysts (including the OECD-AIAS ICTWASS database) conclude that collective bargaining coverage in Italy is 100 per cent. This is problematic, however. Methodologically, it is not consistent with the practice of other countries where there are ways to legally enforce some collectively agreed minimum standards, rather than the whole of a given collective agreement’s provisions, but coverage is calculated only on the basis of the companies that apply the whole agreement, as in some sectors in Norway (Dølvik and Marginson, 2018). In fact, as accepted by the European Commission (2020) in its impact assessment, existing survey data, such as the EU Structure of Earnings Survey or Eurofound’s European Company Survey, suggest that actual coverage in Italy is closer to 80 per cent. More concretely, the judicial route exhibits some problems of weakness and opacity, which expose it to serious risks of wage dumping.

The vulnerabilities of Italian collective bargaining

A judicial route to adequate minimum wages fails to guarantee the same wage standards for equal work (Bavaro and Orlandini, 2021). It makes it impossible to determine ex ante what the legal minimum wage should be in many economic sectors, for four reasons in particular.

First, judges do not consider all economic aspects of national collective agreements in their determination of minimum commensurate wages. This is relevant because in Italian collective bargaining wages are made up of different components, including a variety of bonuses, indexation elements, and deferred payments (such as 13th and 14th monthly salaries, or a severance pay). The 2018 bipartite Factory Pact distinguishes clearly between minimum economic treatment (Trattamento economico minimo, TEM) and total economic treatment (Trattamento economico complessivo), the latter being partially devolved to company and local collective bargaining. As the Factory Pact is still in the process of implementation, it remains to be seen whether in future the courts will consider only the former (TEM) in their deliberations on constitutionally guaranteed wages, even if these may be 20 or 30 per cent lower than the total collectively agreed remuneration.

Second, the courts have not been consistent in their deliberations, which creates broad scope for uncertainty. For example, they have not been consistent in the inclusion of the different wage components mentioned above, using wide discretion in determining which of the components are ‘consolidated’. But also, there is no guarantee that the courts will keep referring to national collective agreements. There have been cases in which courts have allowed lower wages in consideration of specific circumstances, such as economic problems at the enterprise or local economic conditions (Orlandini, 2018).

Third, the fact that minimum wages set by national collective agreements are not prescribed by legislation contributes to non-compliance. The obligation to respect collective agreements derives merely from courts’ interpretation of the Constitution, which in the Italian legal system does not constitute a binding precedent. Wage entitlements can be enforced only through private legal action, and not through the Labour Inspectorate or other enforcement mechanisms. Vulnerable and precarious workers are unlikely to initiate legal action, especially while they are still employed, and trade unions are not entitled to act in court on their behalf on wage matters. Moreover, the obligation is merely a civil law matter, and there are no sanctions for non-compliance (as there is no statutory obligation in the first place). Sanctions can be imposed only in cases of other concomitant legal violations, such as – in particular – illegal gangmasters (caporalato), which can determine the legal offence of worker exploitation. While there is no space for Labour Inspectorates, there is a potential role for social security inspections, given that, by law, social contributions have to be calculated on, at least, the minimum wages set by the relevant national agreement signed by the most representative trade unions and employers’ organisations. While the Labour Inspectorate can then provide workers with documentary evidence to claim their wages, it still only directly controls social contributions payments, and not wages, and in practice it has so far had limited impact.

Fourth, the judicial system of collective bargaining enforcement has been undermined by the multiplication of national collective agreements, signed by a variety of trade unions and employer organisations (Treu, 2019). Both employee and employer representation in Italy are characterised by a plurality of disparate actors, since the 1980s on the trade union side and since the 2000s even more so on the employer side. Legal attempts at regulating representativeness have so far been discouraged by interpretations of Article 39 of the Constitution, which protects trade union freedom, and a bipartite agreement in 2014 to introduce representativeness criteria has not yet been implemented (Recchia, 2017). Almost 1000 national collective agreements have been registered at the tripartite National Council of Economy and Work (CNEL), almost twice as many as there were 10 years ago. Many of them are signed by organisations of very dubious representativeness, but in the absence of statutory legislation, it is left to the courts to select which agreement to enforce, namely the so-called ‘leader agreement’. This may happen in cases in which the employer does not follow any collective agreement, but even more sensitively when the employer follows a national agreement with lower wage minima than another.

As a result, in Italy wage dumping does not occur ‘vertically’ – that is, via derogation through lower-level bargaining – but rather ‘horizontally’, through the application of the most convenient national collective agreement (Lucifora and Vigani, 2021). The domain of an agreement is only defined by the signatories. The courts can limit wage dumping when a company applies a CCNL from a sector that does not correspond to the company’s activity, by applying an outdated norm (Article 2070 of the Civil Code) that was written for Fascist-era corporative contracts. But they have real difficulty in disallowing a sector-relevant collective agreement as signed by non-representative organisations, given the absence of legal criteria. Furthermore, some lower-wage agreements are signed by the largest trade unions themselves, to cover either some subdomains or when negotiating with different employer organisations (for example, with small business organisations or with cooperatives). A notorious case is the ‘multi-service’ CCNL, which can be used by any company to pay a large variety of workers lower wages than those deriving from the CCNL of the company’s main activity sector (for example, logistics).

In some areas, Italian legislation makes reference to the national agreements signed by the ‘comparatively more representative organisations’ in order to set adequate wage standards (such as in the 2007 Law on cooperatives, the 2016 Law on public procurement, the 2019 Law on platform food delivery and the 2020 Law on air transport). These norms can be considered targeted forms of collective agreement extension, but even they do not entirely solve the problem, in the absence of legal definitions of the relevant domains: smaller organisations may claim representativeness in any smaller sub-sector (Centamore, 2020). For instance, in 2020 a new association of food-delivery platforms signed a CCNL with a new trade union organisation (which joined the right-wing confederation UGL), claiming that within the food-delivery sub-sector they were more representative than the organisations that signed the (better) logistics-sector CCNL (Carinci, 2021). In recent years, concerns have been expressed about the possibility of such ‘pirate agreements’ being used to avoid existing ones. Even if from the CNEL collective agreements archive it appears that most workers (97 per cent, but with the reservations about these data that we explain below) are covered by agreements signed by the largest national organisations (on the union side, CGIL, CISL and UIL), the potential loophole constitutes a frequent threat, especially, but not limited to, in emerging sectors and in niche ones.

In this complex and uncertain legal framework, official data such as CNEL’s become debatable, given the unresolved competition between different collective agreements. The OECD’s estimate of 100 per cent coverage considers the courts’ application of Article 36 of the Constitution as a functional equivalent of erga omnes general legal extension, but the suitability of this indicator is increasingly questioned by other analyses. Other sources using the CNEL archive of collective agreements reach estimates of around 93 per cent and surveys around 80 per cent. But the deeper problem, following our analysis, is that surveys are themselves of little use in a situation in which the idea of collective agreement coverage is itself subjective: to what provisions (for example, to what wage components) does it refer to? And to which CCNL? CNEL data are based instead on company social security declarations, which must indicate a CCNL signed by the most representative organisations as reference: but these declarations are a compulsory administrative requirement for the only purpose of determining social contributions, and do not necessarily imply that wages are actually paid according to the agreement indicated in the declaration. A number of national agreements in the CNEL archive appear to cover few employees or even none, which would beg the question of why they were signed and registered in the first place. As evidence of this inconsistency between CNEL data and reality, the collective agreement of Fiat-Stellantis, which has been widely discussed and contested since it replaced the national metalworking agreement for thousands of workers in the Fiat group in 2010, in the CNEL database appears to cover zero employees. The evident lack of consistency between the official data based on social security declarations and actual implementation of collective agreements opens up the potential problem of correct implementation of the Directive with regard to compliance with the minimum wages set out in national sectoral collective agreements.

Ways forward?

While the Italian governments of 2019–2022 had put a national minimum wage in their programmes, the new right-wing executive stemming from the 2022 elections appears to have abandoned such an idea. In November 2022, the Parliament passed a majority motion on the implementation of the EU Directive that recommends the strengthening of collective bargaining instead of a national minimum wage. The motion refers to some of the problems listed above, but its recommendations are vague and refer ambiguously to strengthening company and local bargaining rather than the sectoral kind.

It is possible to envisage two main routes to strengthen adequate minimum wage setting in Italy, a minimalistic and a radical one. The first would involve expanding some solutions that have already been used in legislation on specific sectors, such as in the laws on cooperatives, public procurement, airlines and food delivery. These norms make explicit reference to the national collective agreements signed by the comparatively more representative organisations to set adequate wage standards (although not a formal obligation to apply the whole agreements as such: similar forms of targeted extension exist, for instance, in Norway), and have already been accepted by the Constitutional Court (Judgment no. 51/2015) as compatible with the principle of trade union freedom (Article 39 Constitution). This kind of formulation could be used to establish a general statutory right to the wage standards established by the largest organisations. This would strongly assist the proactive enforcement of the minimum wages set by sectoral collective agreements and reduce the uncertainty that the court route currently entails (the first three problems we listed). But it would still not solve the fourth problem, namely the uncertainty about the scope of application of different collective agreements and the consequent persistent risks of competition between overlapping CCNLs or even of ‘pirate agreements’ in some sub-sectors.

The more ambitious solution, namely to address all four problems, would be to introduce legislation on representativeness, as grounds for declaring the legal extension of collective agreements. France’s experience with the Law on representativeness of 2008 shows that reaching a sufficient consensus for such norms is arduous but not impossible, even in politically divided and adversarial systems. Italy’s social partners have made an important step forwards on agreeing representativeness criteria, and CGIL has made proposals for legislation (namely, in its bill ‘Carta dei diritti universali del lavoro’, adopted in 2015), but the majority of trade unions (CISL and UIL, accounting for more than 50 per cent of all union members) remain opposed to a law on representativeness. Importantly, for such a law to be effective against ‘pirate agreements’ it would need to define the relevant domains of collective agreements in terms of sectors. This constitutes a further difficulty, as the Constitutional Court could strike down such regulations as incompatible with the principle of union freedom from legal obligations (Article 39 Constitution, a norm that does not exist in the French Constitution and has been considered difficult to combine with erga omnes extension). A solution that would make it possible to define domains while respecting trade union freedom would be to leave deliberations on such demarcations to national decision-making by the largest organisations, as proposed by CGIL.

To conclude, while Italy does not need to pass legislation to meet the requirements of the Directive, it should take some action to meet its spirit. In particular, in order to reinforce collective bargaining on adequate minimum wages (Article 4, para 1) and provide transparent data on their functioning (Article 10), it needs clearer, more enforceable norms on the applicability of collective agreements. This need is independent of current debates on whether a national minimum wage should be introduced. If it is not, a better collective bargaining system will be more capable of acting as a ‘functional equivalent’; and if a national minimum wage is introduced in the future, clearer norms – as in the solutions proposed above – will diminish the risk, strongly feared by Italian trade unions, that legal wage regulations might displace collective bargaining and collective representation.