Blurred boundaries, fluid crimes

In its narrowest meaning, caporalato refers to ‘an illegal form of labour intermediation in which the recruiter, a gangmaster called caporale, profits by enrolling and mobilising workers according to the short-term demand of employers’ (Bagnardi Reference Bagnardi and Ledeneva2024, 320). As Perrotta (Reference Perrotta2014) has shown, such brokers have operated around Italy since at least the sixteenth century. Despite profound changes in agricultural production and the development of worker organisations over time, informal brokers continue to be essential today in connecting employers and labourers. Indeed, Perrotta and Raeymaekers (Reference Perrotta and Raeymaekers2023) highlight the functional role of caporalato in contemporary capitalism and trace three interconnected developments since the 1980s that have aggravated the phenomenon: the growing integration of Italy’s food production systems into global markets, state retrenchment from welfare and worker protections, and the growth of foreign labour migration. The authors view caporali as neoliberal brokers who provide flexible and disposable workers, allowing employers and the state to avoid carrying the costs of reproduction of that labour. While not all exploited labourers are migrants, the lack of legal and social protections that many foreign workers experience makes them particularly exposed (UNHRC 2019; Caruso and Corrado Reference Caruso and Corrado2021). Thus, contemporary exploitation mechanisms are entangled with Italy’s restrictive migration regime and the grave situations of precarity and vulnerability it produces (Palumbo Reference Palumbo2024). From this perspective, caporali are not the cause but a symptom of larger dynamics of inequality.

Seeing caporalato as a corollary of contemporary globalised production systems shifts the focus away from considering it a pathology particular to Italy or to agriculture. Indeed, similar practices occur within food production in other advanced economies (Perrotta Reference Perrotta2014; Palumbo, Corrado, and Triandafyllidou Reference Palumbo, Corrado and Triandafyllidou2022; Associazione Terra Reference Terra2021). Moreover, caporalato-like relations have proliferated in other Italian sectors, such as construction and food delivery, logistics, ship-construction and the hotel industry (Omizzolo Reference Omizzolo2021a; Iannuzzi and Sacchetto Reference Iannuzzi and Sacchetto2022; Peruzzi and Piro Reference Peruzzi and Piro2024). Each of these sectors operate uniquely, according to their specific labour requirements and value chains. Nevertheless, they are all marked by the structural tendency towards hiring disposable workers, along a spectrum of more or less in/formal contractual agreements (Peruzzi and Piro Reference Peruzzi and Piro2024). Many businesses exploit liberalised subcontracting systems which enable them to outsource hiring to cooperatives and agencies able to hide underpayment and contractual irregularities within layers of opaque bureaucracy (Franciosi Reference Franciosi2022), thereby eroding wages and workers’ ability to negotiate rights.

In agriculture, these more subtle methods of exploitation, as well as white-collar techniques for avoiding social security contributions (UNHRC 2019), are increasingly being adopted. Yet, here too, exploitative practices vary based on the produce, geographical context and distribution system (Ippolito, Perrotta, and Raeymaekers Reference Ippolito, Perrotta and Raeymaekers2021). Much Italian food production has adopted intensive agri-industrial processes to compete in global markets. Nevertheless, the sector is buyer-driven, dominated by ‘international super buying centres – that is, alliances between the largest distribution groups aimed at obtaining better contractual conditions through collective negotiation with suppliers’ (Palumbo Reference Palumbo2024, 155; see also Caruso and Corrado Reference Caruso and Corrado2021 and Massari Reference Massari, Atzeni, Azzellini, Mezzadri, Moore and Apitzsch2023). Having to submit to the low prices demanded by these actors means that producers’ profit margins are minimal. In some regions, such as Lazio and Puglia, moreover, growers have predominantly remained small-scale, failing to form consortia that would enable them to withstand the aggressive pricing practices of buyers and transportation firms (Associazione Terra, Reference Terra2021). The perishability of produce and widespread under-mechanisation requires farmers to operate under intense time pressure, as links in larger ‘just-in-time value chains’ (Palumbo, Corrado, and Triandafyllidou Reference Palumbo, Corrado and Triandafyllidou2022, 186). A further structural problem concerns the fact that, officially, agricultural wages are set by national collective bargaining agreements and refined at the provincial level to reflect local markets. However, the system incentivises employing workers sporadically, paying for piecework rather than by the hour, and under-declaring the amount worked, while pushing labourers to accept excessively long shifts to maximise earnings (Palumbo Reference Palumbo2024). Although anti-exploitation legislation (Law 199/2016, Art. 10) has introduced a process for aligning agricultural wages to other sectors to fight these downward pressures, in practice this is having limited effects, as it cannot address the problem of undeclared work. Without the massive implementation of labour inspections to monitor payments, contractual stipulations are easily bypassed.

The historical, narrow concept of caporalato has thus taken on fuzzier contours as methods of informal recruitment and employment have proliferated under neoliberalism. Indeed, in agriculture, relations between workers, brokers and employers vary widely. Certainly, the International Labour Organization’s (ILO) indicators of labour exploitation are often present,Footnote ² from excessive hours and inadequate wages to cases of child labour, sexual abuse and debt bondage. Nevertheless, reports that depict workers as passive victims ignore their agency and capacity for resistance, as well as complex relations on the ground. Much research has in fact underlined that distinctions between brokers and workers are often blurred and dynamics more horizontal than is widely acknowledged (Melossi Reference Melossi2021; Donatiello, Martone, and Moiso Reference Donatiello, Martone and Moiso2024). Moreover, workers sometimes choose recruiters, making decisions that negotiate gratitude, community loyalty, self-interest, resentment and resistance (Perrotta Reference Perrotta2014; Poppi and Travagliono Reference Poppi and Travaglino2019, Associazione Terra Reference Terra2021; Massari Reference Massari, Atzeni, Azzellini, Mezzadri, Moore and Apitzsch2023).

While many small-scale farmers justify engaging illegal recruiters and/or exploiting workers due to their own market weakness and tight profit margins, there is evidence that large businesses, with high profits, engage in such practices through rational choice: the financial benefits vastly outweigh risks of being caught or sanctioned (Omizzolo, Reference Omizzolo2021b). Many of these have no ties to traditional organised crime groups. Nevertheless, mafias are also active here. Donatiello, Martone, and Moiso (Reference Donatiello, Martone and Moiso2022; Reference Donatiello, Martone and Moiso2024) show that food supply chains remain attractive to mafias as channels for exerting territorial power through protection/extortion, while facilitating laundering and reinvestment of criminal profits. In mafia-dense areas, such groups have penetrated multiple links in the chain, from harvesting to transportation, logistics, wholesale and retail (Lo Cascio Reference Lo Cascio2018; UNHRC 2019). In particular, EU and national agricultural subsidies make it advantageous to intercept public funds (DIA 2022a; 2022b).

This sector is thus populated by conventional organised crime and by ‘disorganised’ actors (Reuter Reference Reuter1983): opportunistic entrepreneurs who merge legal and illegal processes, operating alone or through short-term, horizontal networks (Bouchard and Morselli Reference Bouchard, Morselli and Paoli2014). While international discussions on organised and white-collar crime recognise that such fluidity is the norm (Holmes Reference Holmes2024), in Italy, much debate still focuses on how organised crime ‘infiltrates’ the legal economy. This terminology arguably perpetuates a mythical distinction between a healthy, legal economic ‘upperworld’ under attack by a criminal ‘underworld’, instead of foregrounding the structural forces that promote law-breaking. Such mafia-centric thinking is particularly evident in the widespread use of the term agromafia. This conflates mafia investments in agribusinesses (La Via Libera 2021), organised criminal – rather than specifically mafia – schemes (Treccani 2018), and generic ‘illegal activities in food supply chains’ (Legambiente 2023). To avoid such conceptual blurring, it is useful to adopt Holmes’ perspective here (Reference Holmes2024, 201): ‘organised crime is primarily about and involves serious criminal activities; the actors engaging in it must also be factored in, but as a secondary consideration’.Footnote ³

This focus on activities is reflected in how Italy’s legislation has evolved. In 2011, following worker protests in Puglia, Article 603-bis was inserted into the Criminal Code, establishing the crime of illegal recruitment/brokerage (intermediazione) and labour exploitation. It was soon criticised, though, for punishing recruiters only when their activity was demonstrably ‘organised’ and coercive (Gaboardi Reference Gaboardi2017). Moreover, it did not sanction employers at all. In 2016, Law 199 – which is still in place – was introduced to reform the framework. This amended Art. 603-bis, establishing clear indicators of exploitation, criminalising exploitative employment as well as brokerage, and eliminating the need to prove criminal organisation to convict.Footnote ⁴ The law’s wording is purposely broad enough to target any sector and type of actor, while also addressing agriculture-specific problems like chronically low employment standards. Thanks to these reforms, investigations and prosecutions have increased in agriculture and beyond, within and outside mafia-dense areas.Footnote ⁵ Most importantly for this study, though, the law extended beyond repression and introduced measures to prevent exploitation and protect victims, setting the basis for the state’s holistic anti-caporalato plan analysed below.

Beyond repression: collaborative governance against crime

Collaborative approaches to fighting crime have proliferated internationally in recent decades. These tend to be community-based policing projects that focus on improving law enforcement agencies’ local legitimacy and efficiency (Li Reference Li2017). Alternatively, they foreground collaborations between state agencies, with minimal citizen engagement (eg. Cels, De Jong, and Groenleer, Reference Cels, De Jong and Groenleern/d). Communities may help define strategies for reducing crime rates, but they are rarely called upon to identify and act on broader social dynamics underpinning crime. Conceptually, though, collaborative governance extends far beyond such policies. It entails building horizontal processes of decision-making and problem-solving among multiple stakeholders (Douglas et al. Reference Douglas, Ansell, Parker, Sørensen, Hart and Torfing2020), with authority partially delegated away from the state. Such collaborations may span all phases of governance: from identifying a problem, to establishing and implementing policies, to evaluating results and rectifying failings (Peters et al. Reference Peters, Pierre, Sørensen and Torfing2022). The extent to which public authorities direct these interactions exists along a spectrum bridging elements of state-centric control with diverging degrees of non-state actor autonomy.

The figures potentially involved are equally variegated. Multiple public agencies may join, as may diverse actors within a single bureaucracy, hoping to dismantle siloed approaches to composite problems (Douglas et al. Reference Douglas, Ansell, Parker, Sørensen, Hart and Torfing2020). Market players may also participate. However, with erosion of faith in the abilities of government and markets to resolve social problems alone, the need to embed civil society actors in governance networks has become widely recognised. The benefits of involving NGOs, citizen groups and unions are numerous: they strengthen the democratic legitimacy and accountability of policies and give voice to groups often excluded from traditional governance mechanisms. Moreover, they mobilise large numbers of people whose commitment may be deeper than that of businesses or public employees.

Nevertheless, Torfing (Reference Torfing2020, 7-8) makes an important distinction between ‘governance in civil society’ and ‘governance with civil society’. In the former, social organisations take responsibility for alleviating problems of vulnerable groups underserved by the state. Crucial though their services are, they rarely have the funding and capacity to resolve multidimensional problems. Yet, in weak welfare systems like Italy (Albertini and Pavolini Reference Albertini, Pavolini, Ascoli and Pavolini2015), this is the primary model through which many social services are delivered, providing what can be considered ‘a cheap (and inefficient) surrogate for a retrenched welfare state’ (Della Porta Reference Della Porta2020, 942). ‘Governance with civil society’ instead connects state agencies, the economy, and civil society actors through collaborations that are relatively long-term and institutionalised (Torfing Reference Torfing2020).

Italy has an expansive civil society presence, including in the fight against crime. The importance of organisations like Addiopizzo and Libera in promoting cultures of legality is widely documented (eg. Lipari and Andrighetto Reference Lipari and Andrighetto2020), and the role of actors that support victims and partner with local government is now recognised internationally (GOCI 2023). Nevertheless, analyses of ‘anti-mafia’ tend to foreground the state’s legislative and repressive actions (La Spina Reference La Spina, Allum, Marinaro and Sciarrone2019). Civil society activities are often treated as a positive – but ancillary – addition to the state’s efforts. In fact, in 2011, Cayli argued that the two spheres remained antagonistic rather than collaborative and stressed the need to connect top-down anti-mafia policies with the civil-society initiatives proliferating at the time. In particular, he called for a binding framework to coherently steer the mix of direct and indirect, macro- and micro-scale approaches. Whether this plea is now being responded to in Italy’s anti-caporalato strategy is the focus of the discussion below.

The analysis is based on two distinct sets of primary data sources. First, I draw on formal policy documents produced in the preparation and implementation of the National Plan, to identify its official priorities and goals, but also its gaps. These texts are then compared to grey literature produced by civil society actors (NGOs and external evaluators), as a first step towards assessing whether the strategy is effectively meeting its goals. Second, I carried out semi-structured interviews with 15 key stakeholders: five members of activist groups or NGOs providing aid to workers; three ethical food producers; two labour unions; two civil servants charged with formulating and implementing policy; and three independent observers monitoring projects. I identified the pool of potential interviewees through the existing documentation: I contacted NGO consortia, ethical producers and research teams listed in the annual reports of the National Plan, as well as ones referred to in grey and scholarly literature. I then interviewed those who gave their availability. Through a process of snowball sampling, these then connected me to other interlocutors. I sought to achieve as broad a spread of perspectives and geographical contexts as possible. Thus, my interviewees came from a large range of regions: Piedmont, Tuscany, Lazio, Campania, Basilicata, Calabria, Puglia and Sicily. The higher presence of Southern actors reflects the fact that NGOs working there on labour exploitation are more numerous, and have been active for longer, than in central and northern areas. I also sought to reflect the perspectives of actors who are involved in the national strategy and those – mostly ethical food producers – who have chosen to remain external due to their critiques of its approach.

The interviews took place between May 2023 and September 2024 and lasted between 45 minutes and two hours each. These were recorded, with interviewees’ informed consent. The structure of my questions, as well as the coding and analysis process, borrowed from the University of Utrecht’s Collaborative Governance Case Database, which provides a questionnaire to organise reports of governance initiatives around core themes from the theoretical literature (Douglas et al. Reference Douglas, Ansell, Parker, Sørensen, Hart and Torfing2020).Footnote ⁶ My account and analysis of the strategy loosely follows its structure in examining the external conditions and institutional context, the initiative’s internal dynamics, and my informants’ perspectives.

The National Action Plan on paper

As we have seen, labour exploitation is driven by intersecting and mutually reinforcing social, legal, economic and logistical forces. Law 199/2016 took a crucial step forward in acknowledging these intertwinements. Although it focused mostly on repression, it incorporated a holistic angle, delegating the Ministries of Labour, Agriculture and Interior to develop a plan for improving seasonal agricultural workers’ conditions through logistical assistance and care initiatives, involving local authorities, employers, unions and third-sector organisations (NGOs). It also called for the expansion of Italy’s Network of Quality Work in Agriculture to promote ethical farming.Footnote ⁷ An annual report on the plan’s implementation was then to be submitted to parliament.

The plan took some years to take shape. First, in 2018, an Inter-Institutional Committee was formed to devise a strategy for preventing exploitation, protecting victims, enforcing legislation, and developing pathways to legal employment and social inclusion. With logistical and financial input from the ILO and European Commission respectively, a multilevel structure emerged, involving 45 entities, organised into a three-year National Action Plan (2020‒2022). This has since been refinanced for three more years.Footnote ⁸ At its core are all the relevant national ministries, connected horizontally.Footnote ⁹ The Labour Inspectorate, police forces, health, transportation and other territorial services are also integrated, as are regional and municipal governments. It is thus a primarily state-centric governance system heavily focused on coordinating public agencies. Nevertheless, it does incorporate civil society actors in two ways. Firstly, key unions, national-level NGOs and agricultural employers’ associations are included in Working Groups responsible for the Plan’s thematic priorities. Secondly, local and regional governments are charged with coordinating and delivering social services to vulnerable workers, many of which are provided by local NGOs financed through national and regional grants.

It was clear from the outset that although many public and non-governmental actors were providing essential services to workers, these were fragmented and uncoordinated, generating duplication in some spheres and gaps in others. Thus, an integrated service delivery system was needed to connect grassroots expertise with public services and law enforcement. Significant advances were made in mapping areas at highest risk of exploitation and the social initiatives locally available (MLPS 2021; 2022; 2023). The Plan’s annual reports offered progressively more nuanced explanations of how macroscopic forces impacted employers and workers and how migrants faced differential vulnerabilities, especially refugees and women. This increasingly fine-grained understanding was both outcome and driver of ever richer collaborative exchanges. While the first annual report retained a state-centric focus, by the conclusion of the three-year project, over 230 stakeholders were involved and over 100 agreements had been signed between public agencies (e.g. with local transport and health services) but also with employers and civil society, including universities, unions and NGOs (MLPS 2023). The Plan’s architecture proved to be sophisticated, therefore, maintaining a coherent structure while also growing organically, drawing in relevant actors as it evolved.

The National Plan echoed scholarly knowledge of the dynamics of labour exploitation and best practices internationally, aiming to target meso-level factors that contribute to workers’ vulnerability: reducing their geographic and social isolation, mitigating mistrust and avoidance of public services, improving mechanisms for legal recruitment, ensuring that law enforcement protects rather than punishes workers, and deterring employers through frequent inspections and sanctions (UNHRC 2019; Melossi Reference Melossi2021; Palumbo, Corrado, and Triandafyllidou Reference Palumbo, Corrado and Triandafyllidou2022). Nevertheless, scholars’ evaluations have been mixed. On one hand, certain initiatives, such as a collaboration between the Labour Inspectorate and the International Organization for Migration, which deploys IOM mediators to communicate with workers and earn their trust, have been applauded (Palumbo Reference Palumbo2024). In purely law enforcement terms, inspections and judicial interventions have increased.Footnote ¹⁰ Meanwhile, some employer organisations have sought to disincentivise exploitation, such as farmers’ consortia that expel members found to illegally recruit workers (Donatiello, Martone, and Moiso Reference Donatiello, Martone and Moiso2024). On the other hand, there has been criticism of Italy’s application of international legal frameworks on trafficking and exploitation. Palumbo (Reference Palumbo2024), for example, shows how legislative gaps have hampered support and empowerment for victims in practice. Moreover, state protections that do exist are bureaucratically cumbersome, only offering short-term residency permits that do not combat the real barriers to entering Italy’s formal labour market. Labour inspections remain too sparse to be a real deterrent, and some meso-level institutions designed to facilitate collaborations have been slow to develop (Palumbo Reference Palumbo2024). Perhaps the Plan’s most glaring weakness has been its failure to promote a unified ethical farming sector. The Network of Quality Work in Agriculture was intended to be a key pillar of the strategy. By 2022, however, only 5,262 farms had joined (representing 3 per cent of farms that employ workers [AIC 2022]) and by January 2024 this had only risen to 6,600 (Macrì and Orsini Reference Macrì and Orsini2024).

Collaborations in practice: perspectives from the field

While the above criticisms are well-founded, my interviews with civil society actors revealed important nuances. If we view the National Plan as an evolving apparatus through which problems and disconnections can be identified and, hopefully, resolved, it has had some successes. For example, interviewees have noted increased coordination between labour inspectors and some public prosecutors’ offices, speeding up workers’ legal protection to address some of the bureaucratic slowness. Such public agency collaborations have primarily emerged at the local level since the problems they tackle are often territorially specific and require that individuals initiate and cultivate personalised partnerships. As one interviewee put it: ‘each situation is different, unique. So you can’t have a template that works everywhere. It all depends on the human material available and whether there are a few individuals who want to make changes and work together’. Indeed, this echoes scholarship which finds that most governance with civil society occurs locally because this is where ‘there is greater proximity between public and private actors, and citizens and their civic organizations can see what is at stake, what they can influence, and how they will benefit from joint solutions’ (Torfing Reference Torfing2020,16). Representatives of two separate NGOs, for example, recounted how their training of local police and trade unionists had greatly improved the sensitivity with which these approached vulnerable workers. One stated: ‘before, officers just cared about the employers and arresting those breaking the law. Now they try to talk to the workers, they realise they have to be trusted, not scare them off. So they ask them what they need and try to connect them to local services.’ In relation to trade unionists, another declared: ‘they didn’t use to pay attention to those who were undocumented or working informally [in nero], because those people can’t join unions. They didn’t see them as their problem. Now they understand that to protect workers you have to also pay attention to their housing, transportation and family situations. Everything is connected.’ A further observer pointed out that ‘there has been a big change in the way unions deal with situations now that formerly exploited labourers have started to work with them as cultural mediators. This means you get real dialogue and the voices of workers start to be heard up the ladder.’

A core area of success has been in building and supporting NGO consortia through which organisations join forces and streamline services. Where possible, these have integrated with local public agencies to create cohesive referral systems to meet victims’ individualised needs. The most advanced examples are the Integrated Social Hubs (Poli sociali integrati) piloted in 36 high-risk locations in the South. Although their effectiveness depends on the existing capacity of local public and NGO services, these have been found pivotal in improving workers’ protections and care (De Angelis Reference De Angelis and Carchedi2023). Another initiative is the creation of an Anticaporalato Helpdesk that operates 24/7 in 14 languages, accessible through multiple channels, which connects workers – many in conditions of grave exploitation – to the nearest services (Nova 2024). Since over half the users have been outside the five southern regions where the Helpdesk operates, discussions are underway to expand it nationally. This makes it just one example of geographic cross-pollinations that the Plan has fostered, with NGOs in the North learning best practices from their Southern counterparts who have been in the sector longer. The grants channelled through the National Plan have been fundamental at a time of otherwise shrinking funding to social infrastructures. In many cases this money has served to offer victims immediate emergency help – to access residency permits, legal support, short-term housing and work – but in others it has enabled them to develop independent businesses or other paths to long-term economic autonomy. For example, one organisation offers some ‘integration budgets’ of around 12,000 euros per individual to invest in financial stability.

These initiatives, like other successful projects financed by the Plan, originated as grassroots innovations. With time, best practices have filtered upwards, influencing decision-making at the regional and national levels. A notable example is the above-mentioned collaboration between labour inspectors and IOM mediators, which was inspired by a local NGO project that first connected police officers with cultural mediators. Thus, various grassroots actors who previously felt that they were working in isolation, disconnected from the state, now have more structured channels for informing policy. As one interviewee stated: ‘we have brought people’s needs to where policies are defined, where decisions are made on how to allocate funds, and how to implement general principles.’

Interviewees, however, also expressed frustrations about the Plan’s priorities and effectiveness. A strong critique is that it addresses the most traditional forms of caporalato while insufficiently tackling white-collar exploitation through fraud and subcontracting mechanisms that violate labour protections. One interviewee, working in the North, underlined: ‘more than half the workers we intercept are documented, they even have contracts. They work for legal cooperatives, but these underpay them or declare fewer hours worked, so later they can’t access unemployment benefits. And the working conditions are still terrible: hours under the sun with no breaks, long travel to the fields and indecent housing.’ Another major challenge is the failure, so far, to create an efficient system to connect workers to employers. While the Plan aims to make Italy’s public employment centres more attractive to both sides and promote digital platforms to match labour supply and demand, this has not occurred. An observer argued: ‘The employment centres have no idea about harvest times, and when farms will need workers. One centre here published job adverts after the harvest was over and the workers had left. They still focus on traditional jobs and on Italian workers and they often can’t handle the bureaucratic problems that foreigners have.’Footnote ¹¹ Some NGOs have consequently developed apps to connect workers to employers in real-time, but these remain localised and fragmented. None have managed to scale upwards and create a unified and trusted system that can replace the word-of-mouth and personal networks through which recruitment currently operates.

The problem of institutional capacity was also raised in relation to many small municipalities that struggle to coordinate anti-exploitation initiatives due to lack of personnel. Although the state has allocated 200 million euros to create decent housing for agricultural workers (MLPS n/d), the requirement that 90 per cent of that money be spent by the end of 2025 was impossible for many to meet. Similarly, the goal of extending transport services to serve labourers’ routes to work has had disparate responses, depending on individual bus companies’ willingness to channel resources accordingly. Thus, the Plan’s objective that the state, together with civil society partners, should replace the speed and flexibility with which caporali meet the interconnected needs of workers remains unfulfilled. Legal labour intermediation mechanisms are inadequate, the ability to travel to work autonomously is limited, access to housing is improving but informal encampments and dilapidated buildings are still widespread. A core paradox is that the Plan cannot fund initiatives targeting undocumented migrants, yet these are often the workers most vulnerable to blackmail and exploitation.

Although NGO networks try to compensate for such gaps in the state’s reach, fragmentation remains the norm and there is a real shortage of actors in some areas. One observer explained: ‘the third sector is not homogeneous, it changes all the time and good initiatives don’t always last. There are often very few actors fighting in a given location, sometimes no more than ten people: a few people in a local union, a small NGO, some police officers. But in other places there are dozens of organisations focusing on caporalato. It depends where you are and whether there are enough motivated people.’ Tensions also exist between organisations that perceive themselves as competitors for public funds. One interviewee argued that many NGOs are progettifici: project mills that continuously adapt initiatives to the available funds, thereby failing to build continuity and expertise. A fundamental problem is that the Plan perpetuates organisations’ dependency on short-term grants whose renewal is never certain, which in turn produces a precarious and underpaid workforce. As one interviewee stated, ‘once the grant ends, we have to wait months or even years to find out if additional funds will be available.’Footnote ¹² Networks similarly tend to disband when funds dry up, undermining the Plan’s intention of fostering stable horizontal collaborations. While some have achieved remarkable continuity despite these circumstances, their members attribute this to their own capacity to diversify funding streams. For many others, though, financial precarity has made it impossible to maintain projects or expand them regionally or nationally.

Vertical collaborations between grassroots actors and public agencies can also be difficult. Most local NGOs do not sit on the Plan’s national Working Groups; rather, they are represented by the lead partner in their consortium. This filtering of perspectives has resulted in weak upward communication and mutual understanding. One representative sitting on a Working Group explained: ‘we are treated like outsiders. The idea of administrative subsidiarity is very recent in Italian policy and many public institutions still struggle with it. They don’t see us as policy actors …, as proper policy stakeholders.’ Some also complain that employers’ organisations are much more vocal than NGOs in these national fora, creating further imbalances. In the absence of a strong institutionalised voice, some grassroots actors seek personalised dialogue with politicians but, here too, relations are often short-term and unstable. One interviewee summarised the overarching situation thus: ‘it is positive that we now have stronger policies against labour exploitation, but it is a shame that the attention devoted to the problem by the system and the politicians is only intermittent.’

Evaluating Italy’s collaborative governance model

The strongest criticism of the anti-caporalato strategy made by all interviewees concerned its inability to resolve the two fundamental engines of agricultural labour exploitation: Italy’s dysfunctional immigration legislation (Law 189/2002, known as the Bossi-Fini) and an agri-food industry that promotes mass production while doing little to support environmentally and socially sustainable farming. One succinctly argued: ‘as long as the state sees foreigners as nothing but cheap labour, with no rights, nothing will change. Politicians don’t care about how the food is produced, what is important is that voters don’t complain about rising prices.’ Among these glaring structural failures, though, interviewees did identify some small positives. While they all agreed that Italy’s immigration regime is hostage to political opportunism, making reform unlikely, some pointed to the significant expansion of entry quotas for 2023‒2025 as an indicator that, as least informally, pressures from employers and activists to improve legal immigration channels are heard.Footnote ¹³ Similarly, Italy’s implementation of two key measures for improving agricultural conditions were strongly championed by the National Plan: firstly, the transposition into national law of EU Directive 2019/633 on unfair trading practices in agricultural and food supply chains (D.lgs. 198/2021). Together with the ‘Agriculture Decree’ (DL 63/2024), these laws ban the sale of food products at prices below production costs, as well as double-down auctions that previously forced farmers to undercut each other to sell to wholesalers and supermarkets. Secondly, Italy was one of the first countries to implement a social conditionality clause in the EU’s 2023‒2027 Common Agricultural Policy, two years before it became mandatory. This causes farmers to risk losing CAP subsidies if caught violating labour rights.Footnote ¹⁴ The Italian government’s vocal support of these mechanisms can be read as a further upward effect of the National Plan, for which these reforms were priorities.

The ethical farmers I interviewed argued, however, that many of their problems remain unaddressed. Over-bureaucratisation continues and they are burdened by rising overhead costs. None see the Network of Quality Work in Agriculture as a solution. One simply stated: ‘the Network is a joke, a failure’. Another explained: ‘it offers no market advantage; there is no logo or labelling system that would make the products recognisable for buyers, and we would still have to sell through the supermarkets that squeeze us. We need a proper alternative distribution network so we can share transport and sales outlets. The people who buy from us are willing to accept our higher prices; they are committed to ethical consumption. But there aren’t many who can afford that.’ Thus, while Italy has a thriving constellation of social cooperatives and alternative food producers (Fonte and Cucco Reference Fonte and Cucco2017), these are largely separate from the Network. As one put it, ‘the state does nothing to support us. If doesn’t care about sustainability.’ Moreover, there is no real system to monitor whether self-proclaimed ethical producers are fully able to avoid all forms of exploitation. Indeed, two external observers referred to ‘ethical’ farms where recruitment and working conditions were worryingly similar to those under caporalato, caused by limited alternative systems for finding workers and employing them legally. One ethical farmer stated that workers are often not interested in staying long-term, listing various employees who abandoned his farm as soon as better opportunities opened in Northern factories. Thus, even where worker rights are respected, the structural problems of finding and retaining employees in such a competitive market, while also maintaining minimal profit, means that only the most socially committed employers persevere. Indeed, one farmer put it bluntly: ‘most employers have enough trouble surviving financially, they really are not interested in thinking about employing workers in better ways.’

Ideally, collaborative governance mechanisms should foster strong interdependence, trust and communication between constituents. The above discussion shows that while horizontal exchanges have improved between public agencies and between grassroots organisations, the system struggles to connect the two. Decision-making remains primarily top-down, dissipating valuable knowledge of those most exposed to the local, shifting conditions affecting exploitation. Civil society voices are heard, but they remain subordinate; they can only influence the system on the state’s terms. There is no space for agonistic involvement (Dean Reference Dean2018) that could alter state actors’ perspectives and priorities. Borrowing from Emerson, Nabatchi and Balogh’s integrative model for analysing collaborative mechanisms, we can conclude that Italy’s National Plan has achieved some interactive processes of ‘discovery, definition, deliberation, and determination’ (Reference Emerson, Nabatchi and Balogh2012, 13), but not to the full degree of governance coproduction it could have done. Similarly, trust, legitimacy, mutual understanding and shared motivation have been hampered by the limited space for grassroots actors’ upward influence. Consequently, while knowledge and resources circulate, real joint action has been sporadic rather than the norm. In terms of the Plan’s impacts so far, it has improved the lives of some exploited workers and will hopefully expand to empower many others. However, aside from the subtle reorientation of some national policies, it has not altered the external structures that produce the racialised inequalities under capitalism that perpetuate workers’ vulnerability (Dines 2022; Palumbo Reference Palumbo2024).

It should be stressed, though, that collaborative governance on this scale is a meso-level process; it cannot be expected to produce major overhauls in structural problems quickly. As a mechanism for ‘steering the economy and society toward collective goals’ (Peters et al. Reference Peters, Pierre, Sørensen and Torfing2022, 8), it has shifted Italy’s responses to agricultural exploitation from disconnected, ad-hoc initiatives towards greater planning and, crucially, recognition that the state is responsible for prevention and protection, beyond repression. Indeed, the Plan’s refinancing for a further three years reflects the understanding that change at a grassroots level takes time and economic continuity. Moreover, co-governance systems should be able to adapt to evolving knowledge. The insistence by many interviewees that the state must fight less overt forms of exploitation has been met by the creation of a National Plan Against Undeclared Work (2023‒2025) as an offshoot of the anti-caporalato strategy.Footnote ¹⁵ While it is too early to evaluate its activities, its inception signals that expertise developed in agriculture can inform other sectors where dynamics of exploitation and white-collar crime are less known.