Collective bargaining in the private sector in Italy primarily takes place at two levels – industry level and company level. However, recently changes to the system have potentially strengthened company-level bargaining, while the number of agreements signed by non-representative bodies has grown.
Collective bargaining in Italy primarily takes place at two levels: at industry level – the most important – and at company or, sometimes, district level. In addition, national level agreements between employers and union have been used to implement EU-level initiatives – such as the agreement on teleworking.
As well as dealing with specific issues, national level agreements between the unions, employers and the government, have also provided a framework for collective bargaining. In July 1993, a key agreement radically reformed the system as it existed up to that point. The 1993 agreement restructured the links between industry and company level bargaining, laid the basis for a new system of workplace representation (see section on workplace representation) and finally ended the system of pay indexation - linking pay to prices - the “scala mobile”. Other national agreements since then have also made major changes.
The main elements of the negotiating structure are as follows.
Industry level negotiations are intended to ensure that pay keeps pace with prices and should set increases that take account of expected inflation. In addition, industry level negotiations deal with a wide range of non-pay issues such as hours and holidays, leave, information rights and work organisation.
Pay negotiations at company level should provide a mechanism for the employees to take account of particular company-level developments, such as improved productivity on the one hand or the risk of job losses on the other. In addition, company level negotiations also deal with changes introduced by the company such as the introduction of new working methods. Union-employer framework agreements and legislative changes have, since 2011, increased the importance of company level bargaining in the regulation of a range of issues.
As well as company-level bargaining it is also possible for this lower level of bargaining to be conducted for several employers on a district or regional basis. This has occurred to a limited extent, particularly in construction, tourism, crafts and agriculture.
Negotiations in the public sector are broadly similar with unions conducting detailed negotiations with a special government agency Aran. However, the government sets the economic parameters for the negotiations, and, following the 2008 financial crisis, it introduced legislation to freeze public sector pay, resulting in eight years without an increase. There are also no local level negotiations in the public sector.
There are no official statistics on the coverage of collective bargaining, but an ETUI study on collective bargaining in Italy, published in 2019, stated that estimates of coverage had been “consistently stable at around 80 per cent of all employees”. However, this figure is for overall coverage at industry level. Bargaining at company level, to improve or complement the industry-level agreements – as foreseen by the July 1993 agreement – is much less common. Recent estimates suggest that only 30-40 percent of the workforce in the private sector is covered by company level agreements.
There is no mechanism for extending collective agreements to employees not directly covered by them. However, the courts will often refer to the minimum wage levels set in the relevant industry-level collective agreement in individual cases where they are asked to judge whether pay conforms to the constitutional requirement for pay to be “commensurate with the quality and quantity of their work.”
The bargaining framework set up in 1993 came under pressure in the mid-noughties, with the employers calling for bargaining to be made more decentralised and for more importance to be given to company-level bargaining. A series of agreements revising the rules have followed. However, because initially the three main confederations adopted differing positions, two important agreements in 2009 were only signed by CISL and UIL on the union side.
The first major outline agreement on the structure of collective bargaining was signed in January 2009 by CISL and UIL, and, subsequently, the smaller confederation, UGL. This was followed, in April 2009, by a more detailed agreement specifying the rules for the new system in the industrial sector.
This agreement, which was also not signed by CGIL:
- changed the length of time that industry-level agreements run, from four years to three years;
- linked pay increases to the forecast European harmonised consumer price index for Italy, excluding energy consumption, and introduced a new wage guarantee element (“elemento di garanzia retributiva”, EGR”), for companies without company-level bargaining; and
- changed the negotiating timetable, prohibiting strikes during the last six months of an agreement and in the month after its expiry.
Despite CGIL’s opposition to the 2009 framework agreement, most industry agreements since then have reflected the key elements of the deal, such as the three year term, and have been signed by the industry federations of all the three main union confederations, CGIL, CISL and UIL.
The main exception was the agreement for the metalworking industry, which was twice (in 2009 for the period 2010-12 and in 2012 for the period 2013-15) only signed by the FIM (the CISL affiliate in the metalworking industry) and UILM (which belongs to UIL) but not by FIOM, the CGIL affiliate in the metalworking industry. However, in 2016 unions from all three confederations signed the next metalworking deal.
The next national agreement with the main employer’s association, Confindustria, in June 2011 was signed by all three main confederations. It set out clear rules for company-level agreements, whose “development and extension” were seen as “a common objective” of all the signatory parties, and permitted company-level agreements to improve or worsen the terms of the industry-level agreement, provided that this possibility has been allowed for in the industry-level agreement itself.
After taking different positions on a productivity document in produced in 2012, all three confederations signed a draft agreement on representativeness, with Confindustria on 31 May 2013. The detailed agreement implementing this draft, which was signed on 10 January 2014 – also by all three main confederations, contained (unlike the draft) a specific section on the respective roles of industry and company bargaining, as well as setting out who was entitled to negotiate and sign agreements. This repeated the position set out in the 2011 agreement that company agreements could modify industry agreements (including worsening them), but only if the industry-level agreement itself permitted this.
The January 2014 agreement also stated that where the industry-level agreement did not contain provisions allowing such modifications, they could still be agreed at company level in the areas of work performance, working hours and work organisation, in order to deal with “crisis situations or where there is significant investment benefitting the company’s economic or employment development.” (This had already been included in the 2011 agreement.)
This agreement was followed in February 2014 by a parallel agreement with the service sector’s employers’ association Confservizi, and identical but separate texts were also signed by the smaller union confederations, UGL, CISAL and CONFSAL.
In 2017 and 2018, further agreements were signed between the unions and the employers on the arrangements for measuring union support, and therefore their right to negotiate and sign collective agreements. Finally, in September 2019, the three main union confederations and Confindustria reached agreement with the state social insurance body INPS and the labour inspectorate, on the collection of union membership data, which will be used together with information on support for unions in workplace elections, to determine unions’ representativeness and their right to negotiate.
However, although these agreements indicate that unions and the major employers’ associations have now developed and agreed a common and detailed set of rules governing bargaining, this does not mean that they are universally applied. This is because they only bind employers and unions who are members of the organisations, and increasingly agreements are signed by organisations outside the normal structures. Between 2008 and 2019, the number of private sector industry-level agreements registered with the official National Council for Economics and Labour (CNEL) increased from 308 to 824.
In November 2017, the CNEL president estimated that only around 300 of these agreements could be considered “regular”. The remining two-thirds were signed between non-representative employers’ association and unions which also were not representative, and set terms and conditions below those in agreements signed by the main confederations. Although the numbers covered by such “pirate” agreements remain relatively low, their existence, in the view of the president of INPS, “ends by damaging workers and the majority of companies which are in a sound position and respectful of the rights of their own employees”.
Alongside the developing relations between unions and employers, set out above, governments have in recent years intervened much more to set the bargaining rules, always to the benefit of greater local flexibility.
In September 2011, the government of Silvio Berlusconi introduced legislation that permitted company agreements not just to agree worse terms than those set by industry agreements, irrespective of what the industry agreements themselves said, but also to undercut the minimum terms set in national legislation on a range of issues.  These included working time, flexible employment contracts, recruitment procedures, work organisation and job classification and the introduction of new technology.
In 2015 in the Jobs Act, the government of Matteo Renzi increased the influence of local bargaining by giving company-level or local area agreements the same rights as industry-level agreements to make changes to legal rules in a range of areas.
This is a contrast to the past, when the rules for collective bargaining were largely determined by the unions and employers themselves and the role of government was limited. However, it now seem that it is only legislation that will enable the national agreements between the main union confederations and the main employers’ associations to be enforced and the pirate agreements to be eliminated. The government of Giuseppe Conte promised to introduce legislation to do this in September 2019, when the agreement with INPS was signed, but, as of March 2020, this had not yet materialised.
Who negotiates and when?
Negotiations at industry level typically involve the employers’ federations and the industrial federations within the major confederations. In most cases, the industrial federations from all three main confederations agree a common platform of demands and sign the final agreement jointly. For a period, this was not the case in the metalworking industry, where in 2009 and 2012 agreements were not signed by the CGIL federation FIOM, the biggest federation involved. But, with the agreement signed in 2016, this ceased to be the case, with unions from all three confederations backing the deal.
Until fairly recently there were no rules on representativeness, governing who was entitled to sign industry-level agreements. However, a series of agreements signed since January 2014 (see above) aim to provide a framework for assessing representativeness and determining which unions have a right to be involved in negotiations at industry level and to sign agreements, although it is important to emphasise that without legal backing – which currently does not exist – these rules do not apply in all cases. (These agreements also cover representativeness at company level – see below.)
Taken together, the agreements set out a new “representativeness” threshold, based on a combination of membership (the percentage of union members in the industry who belong to that particular union) and broader support (the percentage of votes given to that union in elections for workplace union committees which are known as the RSU). Union membership data is to be provided by companies as part of their existing declarations to the social security body INPS, and the election results are to be collected by INPS and the labour inspectorate. The level of representativeness – the representativeness score – is worked out by INPS taking an average of these two percentages.
Under these rules, a union can participate in negotiations if, in the industry concerned, it has a representativeness score of at least 5% and an agreement is valid if it is signed by unions with a representativeness score of at least 50% plus one in the industry. In addition, the agreement must be ratified by a simple majority of the workers involved, using rules set out in the agreement concerned. Once ratified in this way, the agreement becomes binding on both unions and employers.
However, although these rules have been agreed by the main union confederations and the main employers’ associations, they have not yet come into effect as the mechanism for collecting the information necessary to work out unions’ representativeness’ scores was only agreed with INPS in September 2019. They will also not have universal effect in the private sector as employers and unions which do not belong to the signatory organisations, are not obliged to follow the rules. As well as smaller employers signing so-called “pirate” agreements with unions with few members (see above), the major motor manufacturer FCA (formerly FIAT) is also not covered by these rules, as FIAT left Confindustria, the main employers’ organisation, in 2011.
The situation is different in the public sector, where there has been legislation fixing who has the right to bargaining since 1997.This is very similar to the rules now agreed with INPS, and requires a 5% representativeness score to take part in negotiations, and the support of unions with a combined 51% representativeness score for an agreement to be valid. The score is calculated as an average of membership and union support in workplace elections.
At company level it is the elected union committee, the RSU, which normally negotiates, although very often full-time officials from the unions are also involved.
As in the case of industry-level bargaining, the agreements on representativeness, particularly that signed in January 2014, have set out new rules on who can sign the agreements – as well as the composition of the RSU (see section on workplace representation).
These agreements state that, where there is an RSU, a company agreement is valid if approved by a majority of RSU members. In companies, where the union representatives have been appointed directly by the unions – the union body in this structure is known as an RSA – slightly different rules apply. Here, the agreement must be approved by representatives, who together or separately have the support of a majority of union members in the company. In addition, all employees can be required to vote on the agreement if this is called for by either one of the unions involved or 30% of the workforce. For this vote to be valid more than 50% of those eligible to vote must take part and the agreement can be rejected by a simple majority of those voting.
The timetable for negotiations, as set out in the 1993 agreement, was that industry level negotiations on pay should take place every two years, and that on non-pay issues they should take place every four years. However, this changed following an agreement signed by CISL and UIL, but not CGIL, in 2009. Industry agreements now run for three years, covering both pay and conditions issues.
There are often lengthy delays – sometimes of months or years – between the date an agreement runs out and the date the next agreement is signed and there is compensation for this in the final agreement in the form of lump-sum payments. The figures from the national statistics office ISTAT, for December 2019 show that at that point 44.6% of employees were waiting for their agreements to be renewed and that the average length of time these employees were waiting was 20.4 months. The largest group of workers waiting for a new agreement to be negotiated in December 2019 were in the public sector.
Company level negotiations should also take place once every three years.
The subject of the negotiations
As noted above, collective agreements in Italy cover a wide range of issues. At industry level, they set minimum pay rates, which should protect real living standards by matching inflation, but they also regulate issues such as overtime and night rates, hours, holidays, maternity leave, sickness absence, training, health and safety, the use of temporary workers, disciplinary rules, union rights, social benefits (welfare), recruitment and job classification. In addition, some industry-level agreements include provision for the setting up of joint bodies to develop policy on issues, such as equal opportunities or developments the industry concerned.
At company level, the negotiations, among other things, are intended to deal with mechanisms to increase productivity and foster innovation, as well as how the benefits of increased productivity should be distributed. A survey of agreements by ISTAT, covering the period 2012-2013, found that the issues most frequently covered at company level were: fixed pay elements (61.1%); results-based bonuses (58.9%); hours and work organisation (50.7%); occupational training (44.6%); social benefits (welfare) 38.5%; dealing with the crisis (31.9%); types of employment contract (25.3%); relations with the unions and union rights (24.7%); grading (22.8%); and equal opportunities (15.7%).
In addition, the unions at national level have at times been closely involved in broader political issues such as major changes to welfare provision, taxes, and industrial development, particularly in the South of Italy. However, the extent of union involvement in these issues depends on the overall political context and the wishes of the government and is no longer automatic.
Italy does not have a system for setting a legal national minimum wage, although, as already noted, the courts will often refer to the minimum wage levels set in the appropriate industry agreement in individual cases on pay levels. Some political parties support the introduction of a statutory minimum wage, but the union confederations are more sceptical, arguing instead that the minimum rates in industry-level agreements should be enforced as the legal minimum,
 Italy: institutionalisation and resilience in a changing economic and political environment by Roberto Pedersini in Collective bargaining in Europe: towards an endgame, edited by Torsten Müller, Kurt Vandaele and Jeremy Waddington, ETUI, 2019 https://www.etui.org/content/download/36694/368812/file/CB+Vol+II+Chapter+16.pdf (Accessed 25.03.2020)
 A recent study by the Banca d’Italia noted that its own annual surveys of private companies in industry and services with more than 20 employees show that, between 2010 and 2016, around 20% of companies and something less than half of employees had company-level bargaining. I recenti sviluppi delle relazioni industriali in Italia by Francesco D’Amuri e Raffaella Nizzi, Banca d’Italia 2017. https://www.bancaditalia.it/pubblicazioni/qef/2017-0416/QEF_416_17.pdf Another study bythe CNEL and the Italian statistical agency ISTAT found that, in the period 2012-12, 21.2% of companies with more than 10 employees had company-level agreements, but that this was much more widespread among larger companies, with 69.1% of companies employing 500 or more negotiating company level agreements. Progetto CNEL-ISTAT sul tema “Produttività, struttura e performance delle imprese esportatrici, mercato del lavoro e contrattazione integrativa”, December 2013 https://www.istat.it/it/files/2016/03/Report_Cnel_Istat1.pdf
 Accordo quadro: Riforma degli assetti contrattuali, 22 gennaio 2009
 Accordo interconfederale, 15 aprile 2009
 Accordo confederale fra Confindustria e CGIL, CISl e UIL del 28 giugno 2011
 Linee programmatiche per la crescita della produttivita’ e della competitivita’ in Italia, 16 novembre 2012
 Protocollo d'intesa del 31 maggio 2013
 Testo Unico sulla Rappresentanza: Confindustria – Cgil, Cisl e Uil, 10 gennaio 2014
 Convenzione attuativa del Testo Unico sulla rappresentanza, 19 settembre 2019
 XXI Rapporto mercato del lavoro e contrattazione collettiva 2019, CNEL, 2020
 La giungla degli 868 contratti collettivi. Il sole 24ore, 9 November 2017 http://quotidianolavoro.ilsole24ore.com/art/contratti-lavoro/2017-11-08/la-giungla-868-contratti-204538.php?uuid=AEwRtB7C (Accessed 25.03.2020)
 Comunicato stampa, INPS, 19 September 2019 https://www.inps.it/docallegatiNP/InpsComunica/UfficioStampa/comunicatistampa/Lists/ComunicatiStampa/cs190919bis.pdf (Accessed 25.03.2020)
 Testo del decreto-legge 13 agosto 2011, n. 138
 Decreto Legislativo n. 81 del 15 Jube 2015
 See Italy: institutionalisation and resilience in a changing economic and political environment by Roberto Pedersini in Collective bargaining in Europe: towards an endgame, edited by Torsten Müller, Kurt Vandaele and Jeremy Waddington, ETUI, 2019 https://www.etui.org/content/download/36694/368812/file/CB+Vol+II+Chapter+16.pdf (Accessed 25.03.2020)
 Governo, ok a legge sulla rappresentanza sindacale per fissare salario minimo, il Fatto Quotidiano, 19 September 2019 https://www.ilfattoquotidiano.it/2019/09/09/governo-ok-a-legge-sulla-rappresentanza-sindacale-per-fissare-salario-minimo-e-taglio-del-cuneo-tutto-a-vantaggio-dei-lavoratori/5440911/ (Accessed 25.03.2020)
 Decreto Legislativo 3 febbraio 1993, n. 29, modified in 1997 Art.47-bis
 Press release: Contratti collettivi e retribuzioni contrattuali, Diciembre 9 ISTAT, January 2020
 Progetto CNEL-ISTAT sul tema “Produttività, struttura e performance delle imprese esportatrici, mercato del lavoro e contrattazione integrativa”, December 2013 https://www.istat.it/it/files/2016/03/Report_Cnel_Istat1.pdf