On 28 February 2017, Italian police started to evict residents from the ‘Gran Ghetto of Rignano’, a shanty town of self-built shacks arranged around a number of abandoned farm buildings, which over the last fifteen years has provided shelter for migrant labourers employed in the fields on the vast Capitanata plain of northern Puglia. At the height of the tomato harvest in late summer, this settlement has been home to more than two thousand sub– Saharan African migrants, while during the winter its population drops to a few hundred people. In recent years the ghetto has acquired notoriety as Italy’s principal ‘slum’. Over and above the mainstream media’s penchant for superlatives and its eagerness to map similar (but smaller) situations in other parts of the country, the site was considered a cause for national shame, even by the more virulently anti-immigration, right-wing press, while many newspapers regularly pointed to human rights abuses and the possible infiltration of organized crime (which, on this particular occasion, was the legal grounds for the eviction). Indeed, besides the precarious living and sanitary conditions, much was made of the lawlessness that was seen to be rife in the Gran Ghetto. In particular, migrants, we were repeatedly told, were at the mercy of unscrupulous caporali (gangmasters), who personally recruited teams of day labourers, creamed off a portion of their already pitiful wages and extorted further money for transport and food. Yet – and against all apparent logic – these ‘slaves’ continued to return year after year to the Gran Ghetto to seek shelter and employment. And now, in response to the eviction, two hundred labourers even went as far to protest against their removal outside the prefecture in the provincial capital of Foggia. As a headline in the Turin-based newspaper La Stampa declared: “The new slaves live in the shantytown of shame: ‘But we want to stay here’” (La Stampa, 4 March 2017).

In the late evening of Thursday 2 March, with the eviction still not complete, a fire broke out destroying numerous shacks and causing the deaths of two Malian workers who had obstinately continued to treat the ghetto as their home. ² For a few hours on the following morning the episode was the opening news item on the homepages of national broadsheets such as Corriere della Sera and La Repubblica. Despite the lack of incriminating evidence and with fire investigators leaning towards accidental causes, many journalists speculated that the fire had been started deliberately, probably with the intention of clearing away the last stragglers. ³ Expert commentators who had already covered the issue in the past were adamant that the caporali were ultimately at fault. Under the peremptory headline “The gangmaster system is to blame for the death of the Rignano labourers”, Alessandro Leogrande, writing in the left-leaning magazine Internazionale (8 March 2017), insisted that the Gran Ghetto was the consequence and not the cause of this type of labour intermediation and that only the right to vote would enable migrants to be truly heard. Two days later video images surfaced on media websites of migrants apparently laughing as they observed the blaze. Such an act immediately aroused suspicions and operated to transform the figure of victim into that of presumed malefactor. The right-wing newspaper, Il Giornale commented “we cannot see the faces of those laughing as the latest tragedy unfolds” (6 March 2017).

Irrespective of their contrasting political positions, media outlets all tended to structure their reports around indignation at the unacceptable standards of accommodation and compassion for the deceased, which together interconnected with the imperative of restoring order. This was summed up by the governor of the Puglia Region, Michele Emiliano in his declaration to the press: “The tragic death of the two Malian citizens confirms the need to finish the eviction without delay” (La Repubblica, 4 March 2017). Against this dominant humanitarian/securitarian definition of events sat more banal questions about the limited availability of decent agricultural jobs and adequate housing, the absence of any coherent long-term policy of reforms regarding farm labour, and the strategic position of these workers within the wider supply chain. Such points were sometimes briefly mentioned, other times not at all, but they were never at the centre of discussion.

An immediate goal of the eviction had been to move the migrant labourers to two officially recognized camps, even though many migrants refused to relocate because they considered these new abodes too far from their places of work. Something similar had occurred in 2016 in the neighbouring region of Basilicata. After the demolition of the much smaller Ghetto of Boreano situated in the middle of a farming plain ten kilometres north of the small town of Venosa, its former residents were moved, under the direction of the Regional Government’s anti-ghetto Task Force, to a former paper factory in a gully beneath the town that had been hurriedly turned into a dormitory. Responsibility for running the accommodation was assigned to the local Red Cross, which in its own jargon dubbed the building a ‘reception centre for seasonal migrant citizens’ (Rigo and Dines, 2016). ⁴ Indeed, over and above the bizarre title, the centre was somewhat unique: it was the first operative example in Italy of a workers’ shelter run by one of the world’s preeminent humanitarian organizations. ⁵ The head of the operation, met during a visit to the centre in late August 2016, had in fact cut his managerial teeth during the emergency relief programme that followed the 2009 L’Aquila earthquake and later in running a nearby CARA (centro di accoglienza per richiedenti di asilo, reception centre for asylum seekers) that had been set up in 2011 during the government-declared ‘North Africa Emergency’. In interview, the same operator explained that the labourers all voluntarily chose to use the accommodation and the Red Cross provided no further services on the premise that the migrants were there to work – and “those who don’t work, don’t eat”. ⁶ The facilities were minimal: there were four electric hobs, four toilets and four showers for the seventy residents, many of whom had to sleep on metal camp beds without mattresses. Besides the inadequate and rundown facilities, what was also evident was how the Red Cross had assumed the task of overseeing the work force. Visits from outsiders were tightly controlled including those of institutions and trade unions, which had to make formal requests in advance to meet workers on site. Such restrictions were justified by the need to protect migrants from contact with the caporali. The Task Force’s original plan, besides offering alternative accommodation, had been to register the migrants on official employment lists that, theoretically, would have bypassed intermediation. However, in practice, no farmer signed up to the initiative. In fact, all residents met in the centre complained they had worked little since moving to the shelter due to its distance from recruitment points and the absence of transport (the Red Cross only provided a shuttle bus into the centre of town for those who wanted to go to the shops).

Ironically, it was this facility and not the ‘ghetto’ that appeared more to be a space of physical and social segregation. On the contrary, and without wanting to underestimate their deleterious conditions, the various shantytowns of Boreano and Rignano not only offer access to (low-paid) work, but also represent sites of socialization, autonomous organization and mutual aid, where a range of solidarity and activist networks – from language schools to grassroots unions – have been able to build meaningful, if at times contradictory, political relations with residents.

The cases of the Gran Ghetto and the Red Cross shelter serve as entry points for a critical discussion of the ways in which humanitarianism has reshaped both the representation and management of migrant agricultural labour in southern Italy over the last ten years. Drawing on media analysis and ethnographic research of sub-Saharan labourers in the tomato growing districts of Puglia and Basilicata, the article traces the rise of an increasingly dominant public discourse in the mainstream press that sees migrants as ‘victims’ of human rights abuses and predatory gangmasters. While the media has certainly raised public awareness about the often-appalling situations in Italian agriculture, its recourse to a humanitarian perspective has also worked to conceal the centrality of labour relations and the wider question of the agrifood supply chain. At the same time, humanitarianism has also provided the ideological rationale to a series of governmental responses to poor working and living conditions in the region, as demonstrated by the Red Cross centre in Basilicata. Here I want to develop the argument that ‘humanitarian reason’ – to draw on the term coined by Didier Fassin – has increasingly penetrated the management of labour relations, a field generally overlooked in critical studies of humanitarian action, and that this trend is underpinned by a disciplinary logic that is both functional to the regulation of the migrant workforce and the perpetuation of unsustainable forms of intensive agriculture. I certainly do not want to suggest that humanitarianism is the only factor that has determined recent shifts in the representation and governance of migrant agricultural labour, but given its crucial role in moralizing the issue, it provides a compelling lens through which to comprehend broader questions at stake.

In its most essential form, humanitarianism can be understood as a set of governmental and discursive practices geared towards the protection of life, the reduction of suffering and the transmission of compassion. A number of scholars have recently underlined how the dispositif of humanitarianism has assumed an increasingly pivotal role in political action: from legitimizing emergency interventions in exceptional situations such as natural disasters to framing the everyday administration of social and economic exclusion (Douzinas, 2007; Fassin, 2007, 2012; Tiktin, 2011). At the same time, the pursuit for humanitarian goals can often carry deeply paradoxical side effects, such as reproducing unequal power relations or resorting to the use of armed force. According to the French anthropologist Didier Fassin, humanitarianism has reconfigured the social and political sphere of contemporary societies. Alongside the police state and liberal economy, identified by Foucault as the basis of modern politics, Fassin adds the third pillar of ‘humanitarian reason’, the principle under which moral sentiments – both rational and emotional – are incorporated into the government of “threatened and forgotten lives” (2012: 4) that are brought into existence by being protected and revealed.

Humanitarianism should therefore also be understood as a language that emerges in particular historical moments to influence the ways in which politicians, the media, nongovernmental organizations, global institutions – but also sometimes trade unions and social movements – speak about violence and inequality of the world. As Fassin asserts:

we need to understand how this language has become established today as the most likely to generate support among listeners or readers, and to explain why people often prefer to speak about suffering and compassion than about interests or justice, legitimizing actions by declaring them to be humanitarian.” (Fassin 2012: 3)

In one of his case studies, Fassin discusses the ‘humanitarization’ of social policy in France in the late 1980s, which saw a compassionate register emerge and overshadow the traditional emphasis on repressive action, and the subsequent development of damage reduction policies for drug users and the creation of listening centres for unemployed youths. This shift in emphasis coincided with a change in key terminology:

Inequality is replaced by exclusion, domination is transformed into misfortune, injustice is articulated as suffering, violence is expressed in terms of trauma. While the old vocabulary of social critique has certainly not entirely disappeared, the new lexicon of moral sentiments tends to mask it in a process of semantic sedimentation that has perceptible effects both in public action and in individual practices. (ibid.: 6).

A major recent theme in research has been the relationship between humanitarianism and international migration (Pugh, 2004; Walters, 2011; Agier, 2011; Campesi, 2015; Tazzioli, 2016). Much attention has been focused on the humanitarian management of migration flows and how this is intertwined with the securitization of the phenomenon. For example, Giuseppe Campesi examines what he terms the ‘humanitarian confinement’ of asylum seekers in reception centres in Puglia (Campesi 2015). He highlights the ambiguous function of this infrastructure, which, almost always located far from urban settlements, appears to play more of a containing role than a preliminary stage to integration into society. Furthermore, he notes how humanitarian actors (including the Red Cross) involved in the day-to-day running of these centres are also charged with enforcing rules and discipline, in part because the police only intervene in the case of serious breaches of public order. In another example of ‘ambivalent hospitality’, Fassin (2012: 133-157) highlights the significant switch in outcomes of asylum cases that occurred in France during the early 2000s during which a sharp increase in successful applications based on humanitarian and compassionate grounds – especially for people with serious illnesses and female victims of sexual violence – was counterbalanced by an equally sharp decrease in the recognition of refugee status, while all of this took place against a backdrop of a rise in deportations. According to Fassin, the body of the migrant is thus reconfigured by humanitarianism and “now finds its place in a new moral economy that values suffering over labour and compassion more than rights” (ibid.: 87). This formula resonates with the depoliticizing dynamic that lies at the core of humanitarian reason. However, Fassin’s insistence on the subsequent delegitimization and undesirability of the unskilled migrant worker as a result of being situated outside the bounds of humanitarian reason is less convincing and does not hold up to empirical scrutiny. Indeed, this assumption points to a broader lacuna in critical migration scholars’ engagement with the idea of humanitarianism which, while extremely insightful, has rarely moved beyond the external border, the reception system and the procedure of acquiring residence rights. Instead, from the vantage point of migrants employed in southern Italian agriculture, it is clear that humanitarian government continues to operate beyond these preliminary stages by forcibly penetrating also the field of employment. ⁷

Despite humanitarianism’s wide reach, the question of labour has generally been overlooked in the literature, apart from when discussion turns to the professional employees of humanitarian organizations (Fassin, 2012). Underpinning my discussion in this paper is the conviction that critical insight can be gained by inquiring into the relationship between humanitarian government and the socio-economic relations of production. A preliminary question concerns visibility and invisibility. Miriam Ticktin argues that the global fight against human trafficking in the name of humanitarianism has transformed traffickers into “the worst possible injustice” while other “kinds of responses, injustices, forms of subjecthood […] are rendered unrecognizable and unthinkable” (2011: 162). In a not dissimilar way, the humanitarian government of migrant day labour, even when it acknowledges underlying issues such as the continual cost-cutting in the supply chain, ultimately leaves the productive relations unaltered and undisturbed by placing stress on the safeguarding of life and dignity. Just as the humanitarian government of external borders has focused its efforts, at least discursively, on rescuing victims and stamping out trafficking, so the humanitarian government of migrant agricultural labour appears to recognize just two actors: the victims of human rights abuses and their perpetrators, the caporali.

The discussion of the nexus between humanitarianism, migration and agricultural labour builds on case studies of workers employed in intensive agriculture in the Capitanata plain in the province of Foggia in Puglia and the northern reaches of Potenza province in Basilicata. ⁸ The Capitanata, in particular, represents one of the world’s major producers of tomatoes for industrial processing. In contrast to Italy’s other main centre of production – the ParmaPiacenza district in the northern region of Emilia Romagna – where the harvest is entirely mechanized, the majority of tomatoes in Puglia and Basilicata are picked by hand (Perrotta, 2017). At the height of harvest between August and September, approximately 15,000 migrants are employed in the Capitanata, about a quarter of who are from sub-Saharan Africa and the rest from Eastern Europe, while around 2,000 are employed in Basilicata, the majority originating from Burkina Faso.

The goal of this paper, as I have already made clear, is not to offer an in-depth analysis of these two contiguous tomato-growing districts. While they each possess their own agronomical and historical specificities (just like other farming regions across the Italian South), some general points about migrant labour in southern Italian agriculture need to be made in order to better frame the discussion. First, it is important to stress that the arrival of migrant workers from the 1980s onwards ⁹ coincided with the industrial restructuring of Italian agriculture and the transformation of the supply chain (Corrado, De Castro, Perrotta, 2017). Particularly over the last decade, major food corporations and supermarket chains have assumed almost total power in the fixing of product prices, while labour often represents the only cost over which medium and small producers have some margin of control (Garrappa, 2017; Perrotta, 2017). The continual push to reduce production costs is thus a key reason for the renewed resort to the caporalato in some areas, the historical form of labour intermediation that during the 1970s and 1980s had appeared to be in terminal decline (Avallone, 2017). Moreover, the replacement of local day labourers with migrants in certain sectors, especially in the tomato and citrus fruit industries, coincided with the reform of the agricultural employment system, which essentially consisted in the removal of all forms of public mediation between supply and demand. Moreover, this reform took place in a sector traditionally characterized by the prevalence of locally agreed contracts and hence exacerbated the fragmentation of regulatory frameworks.

Work conditions are notably poor in numerous sectors. The low official rates of pay (which amount to no more than 40 euros a day) have led to a tendency among migrants to opt for piecework. In addition, the common on-call contract has enabled the proliferation of irregularities, such as the incomplete registration of the total number of days worked, which in turn makes it difficult for migrants who possess formal contractual agreements to renew their work permits. In fact, it is not uncommon for migrant agricultural labourers to revert to other means, such as acquiring a contract for domestic work, in order to maintain their residence permit. Public discourse typically conflates these legally established conditions with the local gangmaster system. Certainly, the caporali are at times violent, dishonest and profiteering but they are also themselves often migrants and former labourers who establish relationships based on mutual trust with fellow nationals. Moreover, poor working conditions can be found in sectors where there are no caporali, such as in the greenhouse horticulture in southeast Sicily (Palumbo and Sciurba, 2015) or in the vineyards of southern Tuscany where an analogous form of intermediation is legally carried out by agencies offering ‘third-party’ services (Oliveri, 2015).

The composition of the migrant labour force has undergone a number of significant transformations in the last two decades. A farmer-cum-activist from Palazzo San Gervasio in Basilicata noted how the entry of migrant labour into agriculture had hurled formerly isolated villages in the southern Italian interior “onto the frontline of history”. ¹⁰ This suggestive expression points to how agricultural day labour is periodically influenced by wider political and economic changes at the international level. So, in the specific case of northeast Basilicata, Albanian workers arrived in the early 1990s after the end of the communist regime, followed by North Africans, who were in turn replaced by nationals of Central and Eastern European countries who were able to enter Italy without a visa after they became citizens of the European Union, after 2004 in the case of Poles, and 2007 in the case of Romanians an Bulgarians. The 2011 uprisings in North Africa led to the arrival of Tunisian workers, albeit for a brief period, while the most recent international crises have seen a growing presence of workers from subSaharan Africa who have joined the majority Burkinabé community.

While little more than a decade ago research revealed that over 60% of workers employed in the southern Italian tomato sector were undocumented and 96% were employed without a contract (Medici Senza Frontiere, 2005), the situation has since markedly changed. Today the majority of workers, as already noted, are from Romania and Bulgaria, and are therefore European Union citizens. Recent reports have also indicated that African migrants, frequently the focus of media attention despite their smaller numbers, are for the most part authorized to reside in Italy, even if their documents are temporary and have not been issued for work purposes (MEDU, 2015; Rigo, 2015; FilieraSporca, 2016). Many African workers are beneficiaries of subsidiary and humanitarian protection, having received these documents during the ‘North African emergency’ between 2011 and 2013, and are employed alongside a recently increasing number of asylum seekers, a smaller but not insignificant number of migrants who have lost their factory jobs in northern Italy, and, in a few cases, second generation youths for whom seasonal agricultural labour represents a source of additional income. Most workers now also possess, at least on paper, some form of job contract, although this invariably does not protect them from mistreatment and illicit practices (such as the nonregistration of work days on pay slips), while the obligation on the part of employers to provide adequate accommodation and transport, as we have seen, is rarely fulfilled.

Finally, it is important to note that although migrants once considered agricultural employment a temporary source of income, which was often undertaken as they awaited the necessary documents to move on to look for more stable and better-paid industrial work in the Italian North, today it has increasingly become a permanent, albeit seasonal occupation. In particular, sub-Saharan migrants are prone to move between regions following the cycles of agricultural production, from citrus fruit in Calabria in the winter, to potatoes in eastern Sicily in the spring, to tomatoes in Foggia and Basilicata during the summer and autumn months. Eastern European workers, with their greater transnational mobility, are divided between those employed on a year-long basis, and therefore permanently settled and sometimes with their families, and seasonal labourers who, precisely because they spend shorter periods in the fields, are able to endure harsher conditions and rates of pay that permanently resident migrants, including Africans, refuse to accept.

Although migrants have composed a significant section of the agricultural labour force in Italy since the 1980s, their presence has only recently become a recurrent source of public interest. Prior to the mid–2000s, articles about migrants in agriculture were relatively scarce and cursory: migrant agricultural labour was usually a secondary concern mentioned in relation to a prioritized news topic, such as illegal migration or crimes committed by foreigners. A key exception happened for a few summers in the early 1990s, when the mainstream press reported on tensions between local Italians and migrants from Africa during the tomato– picking season on the Capitanata plain. However, the accent on migrants’ participation in the labour market, which until the 1980s had been a central feature in discussions about immigration (Sciortino and Colombo, 2004: 98-101), hereafter practically disappeared as mainstream media coverage of immigration became increasingly politicized around legislative reforms and security debates (Dal Lago, 1999).

The origins of current, national media interest in migrant agricultural labour can be traced to a famous 2006 investigation in the weekly magazine L’Espresso, in which well-known undercover journalist Fabrizio Gatti, masquerading as a (white) South African labourer, exposed the atrocious working and living conditions in the tomato fields around Foggia (Gatti, 2006). As already made explicit in the title, “I was a Slave in Puglia”, Gatti frames his report around the highly emotive trope of slavery that had largely been absent from previous accounts of migrant labour. The spotlight also falls on the figure of the migrant caporale who docks wages for transport, food, and water, demands sexual services from female workers, and inflicts violence upon pickers for the paltriest of reasons. At the same time, the caporale reassures his squad of undocumented migrants that they are safe from police raids by telling them that their fields are controlled by the mafia. Slavery, gangmasters, and organized crime are all signifiers that exceed “acceptable” forms of exploitation, resonate with timeworn ideas of a backward, violent, and exceptional Italian South, and ultimately seek to arouse the reader’s emotions. Reference to any broader context is largely confined to a few final remarks on the national disparities in tomato prices at production, processing, and retail stages. Slave labour acquires representational currency as the author circumvents the complexities of the global capitalist agricultural system and dwells instead upon a humanitarian discourse: “There is not a single field that meets the regulations for seasonal labour. This is not simply a case of unfair competition in the European Union: it is here…that the most serious abuses against human rights are tolerated” (Gatti, 2006).

Since Gatti’s 2006 report, slavery and the caporalato have increasingly become the key motifs with which Italy’s principal newspapers La Repubblica and Corriere della Sera (but also smaller anti–establishment and leftist dailies like Il Fatto Quotidiano and Il Manifesto) have addressed the issue of migrant agricultural labour. The two themes are typically accorded priority through sensational and exaggerated titles such as “Twelve Hours in the Fields Without Pay: The 100,000 Prisoners of Gangmasters” (La Repubblica, 2 April 2014) or “Gangmasters and Mafia: 700,000 Slaves in Italian Agriculture” (Il Fatto Quotidiano, 10 December 2012).

The recurrent nature of dramatic events – the ghettoes, lethal fires, work-related deaths (that have included Italians and migrants alike), the protests and riots – have offered a steady stream of news and have consolidated a particular humanitarian-inflected perspective on affairs. As in Gatti’s report, migrant workers have generally been depicted as desperate, submissive, and lacking in collective agency. Even when they do organize, they are usually seen as doing so from the position of slaves. Hence, a 2013 report about a legal suit against farmers is entitled: “Revolt of the Tomato Slaves Against the Bosses” (La Repubblica, July 17, 2013). In other moments of conflict, such as the 2010 revolt of Rosarno in Calabria, migrants have been perceived as victims or perpetrators of violence, but rarely as workers making political claims.

From the earlier discussion of migrant workers’ accommodation and the examples reported here, it is clear that the categories and interpretive frames used by the mainstream media to define the issues at stake in agricultural work mark a political and rhetorical break from the discourses about clandestini (illegal migrants) and their causal links with public sentiments of insecurity that had dominated the media’s coverage of the negative aspects of immigration until the mid-2000s. However, this does make these ‘new’ categories and frames less misleading or problematic. First and foremost, the recurrent allusion to slavery does not capture the variety of complex situations in southern agriculture that are mostly characterized by a grey area between regularity and irregularity, both in regards to employment and legal status. Moreover, slavery depoliticizes the question of migrant labour, by absolving the responsibilities of the state and removing the calculated choices made by migrants when deciding to work in the fields (O’Connell Davidson, 2010). Furthermore, the figure of the caporale, as previously noted, is far more ambivalent than usually made out: recent social research on southern Italian agriculture has in fact stressed that in order to operate, most caporali need to maintain good social relations and levels of trust with their respective national and ethnic-based networks (Perrotta and Sacchetto, 2014: 81-82). Likewise, settlements such as the Gran Ghetto, despite their squalid state, provide migrants with a greater degree of sociality, security – and indeed visibility – than, for instance, abandoned farmhouses in isolated areas (ibid.: 79).

By focusing on gangmasters and slums, other key players in the production process are invariably erased from the picture. Indeed, located only a few hundred yards away from the Gran Ghetto is the headquarters of the Caccavelli group, one of the Capitanata’s largest agro– businesses, which claims on its website to be a “nature specialist” committed to organic production and to campaigning against counterfeit imports from abroad (particularly China), but unsurprisingly does not mention its use of migrant labour. This and other local companies rarely enter into the press’ discussions about the tomato harvest. Again, the emphasis on the caporale tends to conceal the industrial and global dimensions of the sector.

It is important to note that a concern for the protection of life and compassion for the victims of violence had previously existed in the Italian media’s coverage of immigration, but that this was intermittent and typically connected with exceptional moments such as racist attacks. It was only from the mid-2000s that humanitarianism emerged as a more permanent discursive frame with respect to certain news topics such as boat crossings and migrant agricultural labour. Moreover, the increasing imprint of a humanitarian logic did not replace previous interpretive frames but rather complemented and sometimes reinforced pre-existing discourses, particularly those of security and order. Responding to Fassin’s call to historicize humanitarianism, a number of processes can be considered instrumental in propelling this particular discourse to the forefront of the public sphere. Among these were: the increase in boat crossings and related deaths in the Mediterranean Sea, initially in the southern Adriatic during the late 1990s and then in the Straits of Sicily from the mid-2000s onwards; the establishment of reception centres for asylum seekers in 2004 (Campesi, 2015), which complemented the existing administrative detention system and led to a blurring between detention and hospitality; the emasculation (although not the disappearance) of the political insistence on the links between immigration and urban insecurity that had reached its height in the late 1990s and early 2000s, thanks in part to the passing of the centre-right Bossi-Fini law in 2002 that had the temporary effect of exhausting the more inflammatory rhetoric in public debates; the contemporaneous pressure both from outside and within the press to redress journalists’ sloppy use of dubious terminology such as ‘clandestino’ ¹¹ ; and the impact of the economic crisis after 2008 that heightened attention to an increasing sense of vulnerability across Italian society.

The recent representations of sub-Saharan migrants in southern Italian agriculture thus point to a broader shift in public discourses about migration. Until the mid-2000s, depictions of migrant agricultural labour were largely structured around the figure of the undocumented migrant, who was either regularizable or deportable but was nevertheless considered a constituent member of a workforce. Agricultural work was typically considered in public debates to be a temporary occupation and those migrants who managed to remain in Italy would eventually move north in search of factory work. Today, allusions about labour relations are overshadowed by images of migrants as victims of slavery and calls to tackle human rights abuses. This switch in representation runs parallel to the increasing use of the humanitarian paradigm in migration and border management. Indeed, over the last few years, it has been the fields of the South and southern maritime borders such as Lampedusa – and not the metropolitan centres of Italy – that have provided a centre stage for the weaving together of humanitarian and security discourses about migration. In other words, the tomato industry of Puglia and Basilicata, the citrus fruit district of Calabria and greenhouse horticulture in Sicily and Campania have become, over and above the extreme situations of marginalization to which they have been associated, a testing ground for a shift in political perspectives on migration.

The argument that governmental and non-governmental institutions have increasingly adopted a humanitarian approach to confronting the question of migrant labour in intensive agriculture is not meant to simply highlight the incorporation of categories such as ‘slavery’ and ‘human rights’ into policy language. Rather, it also concerns the process by which certain critical issues are selected and subsequently addressed as emergencies, and to how these critical issues and the responses to them are both implicated in the government of borders and migration flows. This process is, at the same time, extremely contradictory: it often misidentifies symptoms as structural causes; it concentrates on combating certain figures (caporali and the ‘mafia’) while leaving others (e.g. food corporations) out of the picture; and it treats regular, irregular and EU migrants under the same humanitarian umbrella, regardless of their different situated demands.

The humanitarian government of migrant agricultural labour often appears to be a reactive strategy, in the sense that it tends to be called into action following exceptional events or at the height of harvests when the number of labourers increases and the injustices and infrastructural inadequacies are at their most conspicuous. However, it has also increasingly come to predetermine the content of legislation and the political questions at stake. As noted, the significant changes in agricultural production in recent decades have not been accompanied by comprehensive labour reforms. In particular, the resort to migrant labour on the part of farmers and producers in order to ride out such changes has not been addressed by specific labour or social policies nor has it aroused interest within the field of labour law (Rigo, 2015), but has instead been regulated through the management of migration flows.

The ‘humanitarian turn of events’ has not only concerned work in agriculture but also the government of migrant labour more generally. Before 2011 Italy’s labour demand was formally managed through an annual quota system – the so-called ‘decreto flussi’ (decree on flows) – that regulated the entry of third-country national workers. Each year the government also issued specific quotas for seasonal work in various sectors including agriculture, although as the system persistently underestimated numbers labour needs were always met in part by irregular workers. Since 2011, largely as a result of the economic crisis and the North African emergency, the Italian government has effectively blocked this system, maintaining limited access for only those who attend training programmes in their countries of origin and highskilled migrants and cutting almost 50% of the quota for seasonal workers. In the case of the tomato sector in Capitanata, the potential shortage in labour has been partly compensated by the recruitment of EU seasonal workers, but during peak harvest time and in the neighbouring region of Basilicata, third-country nationals continue to comprise a crucial segment of the workforce. If the majority of these workers are recent arrivals with some form of legal status, a legitimate question thus arises as to what mechanism has taken the place of the previous quota system for (cheap) labour. Without wanting to suggest a calculated policy on the part of the State, one clear explanation is to be found in the recent Mediterranean ‘refugee crisis’ and, prior to this, the North African Emergency. Indeed, it is possible to identify a degree of overlap in terms of statistics: the highest number of asylum seekers arriving at Italy’s maritime borders was 181,000 in 2016, which is not much more than the 158,000 workers foreseen in the final ordinary decreto flussi for seasonal and non-seasonal workers s in 2011 (Rigo and Dines, 2017). ¹²

Of course, this is an arbitrary comparison of a very different set of figures. Nevertheless, their superimposition offers an alternative perspective on the so-called ‘migration crisis’ insofar as the maritime arrivals since 2011 do not compare with the much higher annual number of migrants that Italy absorbed in the past, and moreover it illustrates that the formal block on work quotas has certainly not led to a drop in the request for low-cost migrant labour or to the removal of recruitment channels. Rather, there has been a decline in the mechanisms that stabilize the position of migrants as bearers of rights, which – however questionable and difficult to put into practice – had until recently formally justified migration policy. One could object that the current composition of migrant labour in agriculture simply reflects the crisis that has reshaped Europe’s borders and that once the current (albeit apparently enduring) period of crisis comes to an end, the distinction between ‘regular’ and ‘illegal’ migration will return once again to the fold as the fundamental principle for managing public order. However, such a premise would underestimate the constitutive force of the humanitarian turn within the government of migration.

A key impact of humanitarian government needs to be understood in the nexus between migrant workers’ accommodation, the reception system and access to residence rights. In his study of the asylum seeker reception infrastructure in Puglia, Giuseppe Campesi (2014, 2015) discusses the centres’ porosity with regards to the informal economy that takes place on their immediate outside. In particular, the CARA of Borgo Mezzanone in the Capitanata has become a night shelter for workers employed in the local countryside during the peak season, who are usually able to enter and exit the centre through various holes in the external fencing. The renovation of the CARA in 2010 led to the dumping of the old prefabricated buildings on the adjacent airstrip, which were subsequently occupied by migrants and asylum seekers who continued to use the facilities of the official centre. According to Campesi (2014), the reception centre and the informal settlement have become indistinguishable, as the whole site has become a main attraction for agricultural labourers looking for accommodation in the area.

It should come as no surprise, then, that the institutional solutions proposed for seasonal workers living in the various ‘ghettoes’ in southern Italian agricultural regions mobilize the same humanitarian actors and methods that are deployed to respond to the needs of people displaced by wars or natural disasters. Examples include the 2014 “Capo free Ghetto off” operation coordinated and funded by the Puglia regional government, which attempted to move the workers from the Gran Ghetto to tent cities set up by the Civil Protection, and the already cited Red Cross shelter in Basilicata for labourers who had formerly resided in the Ghetto of Boreano. Meanwhile, in the case of the Gran Ghetto itself, the basic needs of its inhabitants had been granted by emergency public funds that, until recently, were renewed every year. This saw the regional government supply drinking water and subcontract the provision of portable toilets to a private company. Over and above the good intentions of eradicating or alleviating insalubrious living conditions, we are faced with the pressing question: why does the management of a fundamentally important, permanent albeit seasonal, segment of the agricultural labour force treat workers as if they were refugees and agricultural production as if it were an unpredictable natural event?

This question is by no means rhetorical but hints at a series of legal and practical implications that call for serious consideration. The corollary of endorsing policies that aim to tackle a situation constructed as being in a perennial state of emergency is that rules and access to rights are continually negotiated rather than established once and for all. First and foremost, the lack of an official address translates into often not being in a position to access social services or to renew a resident permit. So, for instance, the fact that the Gran Ghetto and many other settlements are not formally recognized has long-term repercussions for the lives of the residents. Their prospective ‘rescue’ by institutions offers them a stark choice between remaining in invisible informality or entrusting themselves to humanitarian visibility (but usually with reduced opportunities for work). The problem of the lack of in-situ residence applies to both sub-Saharan and EU labourers, even though the latter enjoy greater mobility. In addition, asylum seekers and beneficiaries of international or humanitarian protection are almost always registered at the headquarters of the voluntary associations or reception centres where they first submit their application, with the result that they usually move to live and work in municipalities that are different from the ones where they officially reside, which is certainly advantageous for those administrations that host large numbers of labourers and would have welfare obligations to them if they were residents. These migrants thus have to intermittently travel to other parts of Italy in order to renew their resident permits. Commentators have rightly indicated how the residence registration procedure has de facto turned into a security practice for the control of internal mobility, which needs to be considered on the same level as the government of external borders (Ronchetti 2012; Gargiulo 2013).

The difficulty in renewing residence is acutely apparent in the CARA, where formal and informal residents both negotiate the length of their stay with the administrators of the centre. The CARA also hosts a number of informal residents who have overstayed the maximum sixmonth period but who are allowed to stay by the third-sector and non-governmental organizations that administer the centre because they have no other alternative. In addition, there are individuals – including those working in agriculture – who return to the CARA for the specific purpose of renewing their documents because its address remains the official domicile in procedures regarding legal status. The upshot of this situation is that humanitarian operators are assigned wide-ranging powers over the lives of migrants that can include administratively revoking reception measures in the event of a person’s repeated breach of rules. ¹³

The pre-eminence accorded to human rights violations over labour relations that was noted in the opening discussion of the Gran Ghetto’s eviction is also confirmed by institutional measures that have largely failed to effectively confront the reality of agricultural production. A case in point is the 2012 transposition of a EU directive into Italian law that provided sanctions against employers of illegally staying migrants (Directive 2009/52/CE), which was initially hailed as an innovative instrument in the fight against labour exploitation because it introduced the possibility of issuing a resident permit to those migrant workers who, subjected to a condition of particular exploitation at work, decided to denounce their employers. However to date such arrangements, introduced as incentives for reporting irregular labour practices, have had almost no effect (only 25 permits were issued for all categories of employment between the start of 2013 and February 2016). First, one needs to bear in mind that the EU directive is primarily conceived as a measure to counter illegal immigration and so only regards those who possess no document whatsoever. Therefore, it can only be applied to a very small proportion of the agricultural workforce, given that most migrants have some form of legal status. Second, the incentives for migrants to disclose their illegal working conditions are envisaged only when they are subject to extreme exploitation, which, as defined by the penal code, is very difficult to legally prove. In other words, the underlying paradox is that employment rights are recognized only when migrants are perceived as victims of semislavery conditions and not simply as workers. ¹⁴

The unequal relations that are reproduced by humanitarian government do not therefore simply address the condition of suffering but extend to questions of discipline and production. In humanitarian government, subordination and discipline are the ordinary means of accessing the rights to which one is entitled. These rights are not guaranteed once and for all, but need to be continually renegotiated with public and private subjects, whether this is the voluntary organization that offers to provide a residential address for the renewal of permits or the management of the CARA which closes an eye to those who reside beyond the terms allowed by law because they have nowhere else to go.

Reflecting on the functional coupling between humanitarianism and securitization in external border management, William Walters argues that the former ‘compensat[es] for the social violence embodied in the regime of migration control’ (Walters 2011: 139). From the case study discussed in this paper, it would appear that humanitarianism at least attempts to compensate for the economic violence that is exerted upon migrant day labourers as a result of the continual push to reduce production costs in the agricultural supply chain.

If the securitization paradigm implemented to fight ‘illegal’ migration establishes a distinction between regular migrants, as entitled to a set of rights, and ‘clandestini’, as detainable and deportable, the humanitarian paradigm is far more nebulous. As Michel Agier has suggestively posited, the humanitarian realm is governed through “striking with one hand, [while] healing with the other” (2010: 29). Whether on the external border or in the tomato field of the Italian South, the humanitarian paradigm finds its justification in conditions discursively constructed as exceptional, be these the extraordinary migratory flows or the inhumane exploitation faced by migrant harvesters.

As Fassin notes, humanitarian reason is constituted by an asymmetrical relationship of compassion: “When compassion is exercised in the public space, it is therefore always directed from above to below, from the more powerful to the weaker, the more fragile, the more vulnerable – those who can generally be constituted as victims of an overwhelming fate” (Fassin, 2012: 4). Nevertheless, those at the receiving end of humanitarian attention can sometimes tactically appropriate its language and technologies for their own ends. Many migrants prefer to maintain their humanitarian or international protection status rather than convert their documents into work permits. The latter are subject to stricter requirements that can easily result in migrant workers finding themselves in situations of irregularity regarding their residence. If, on the one hand, humanitarianism tends to depoliticize the demands of migrants, on the other hand, there are opportunities for migrants within its interstices to renegotiate access to forms of citizenship. As such, rhetorical discourses about migrants as victims of slavery need to be critically deconstructed, but so do superficial critiques that equate the notion of victimhood with the removal of politics. As Partha Chatterjee (2004) argues, political society is not always engaged in processes of subjectification that are immediately comprehensible or acceptable for a liberal discourse that perceives rights as a form of emancipation. In other words, if a postcolonial critique is to have purchase on such issues, it not only needs to address the contingent nature of humanitarianism, but it must also take into consideration the ways in which migrants reveal the contradictory and unsettled grounds upon which humanitarianism itself is constructed.

Behind the dominant public and policy narratives about a humanitarian emergency in the tomato fields and citrus groves of southern Italy, diverse discourses are elaborated, alternative forms of agricultural production that bypass the supply chain are experimented and instances of political action are organized by migrants and their allies, such as the successful blockade of a tomato processing plant outside Foggia in late August 2016. ¹⁵ The long-term prospects for improving the overall situation are somewhat tempered by Domenico Perrotta’s pessimistic diagnosis (2017) that if migrant tomato harvesters in the Capitanata were able to successfully secure better working conditions, this would result in their almost immediate replacement with harvesting machines, just as what occurred in California in the 1960s and in Emilia Romagna in the 1980s. ¹⁶ In the meantime, humanitarian government helps to keep this contradictory tension in check by playing its part to prolong the inequitable relations of production.