A year into a pandemic that has put them under ever more strain, yesterday, Amazon workers held a twenty-four-hour strike across Italy. In the first nationwide strike in the company’s history, workers mounted picket lines to protest exhausting work rates, despotic management-by-algorithm, and the company’s lack of accountability to its hires. The strike day was particularly historic because it involved all Amazon logistics workers, from warehouse employees to delivery drivers.
An obvious key focus for the strike was Amazon’s large distribution centers (“fulfillment centers”) where thousands of goods are stocked, picked, and packed, yet the strike extended to the mid-range sortation centers (where boxes are dispatched) and the small “last mile” delivery stations. Decisively, it also included the drivers, who are outsourced and not recognized as Amazon employees, even though they work under the direct control of its algorithms.
Claiming 75 percent participation, the one-day strike thus represented a historic moment for the labor movement — and for Amazon itself. Yet the struggle also needs to be extended, including to the international level, if it is to really hit Amazon’s logistics operations and push back against the company’s most exploitative practices.
Calling the Strike
The strike was called on March 10, after the sudden breaking off of negotiations between Amazon Italia Logistics (a subsidiary running operations in seven Italian fulfillment centers) and the logistics branch of the CGIL, CISL, and UIL union confederations. After two meetings back in January, unions expressed satisfaction that a discussion was underway. Yet the company made no concrete commitments on their specific demands, centered on a company-level collective agreement on working conditions, health and safety, work intensity, schedules, bonuses, and meal vouchers.
The company could not have refused all discussion. In Italy, trade unions still enjoy relatively strong institutional power: total membership is among the highest in Europe and unions are still influential in policy making. But, consistent with its classic strategy, Amazon worked to buy time rather than respond to union demands. Despite unions’ optimism that they could negotiate traditional industrial relations in the firm, social dialogue turned out to be a dialogue of the deaf.
The negotiations blew up two weeks ago at a meeting where the company refused to recognize its social responsibility toward subcontracted drivers. It issued a statement insisting that “for deliveries to customers, Amazon Logistics uses third-party suppliers. Therefore, we believe that the correct interlocutors are the suppliers of delivery services, as well as the business associations that represent them.” The three unions blamed the company for the collapsednegotiations and declared a national strike — involving not only drivers but the entire national distribution network.
In the past, the different working conditions in different parts of the logistical chain, as well as the different levels of union organization, had prevented any such nationwide strike. The earliest organizing took place at the Piacenza fulfillment center, opened in 2011. It took five years to unionize the first group of workers there, and in 2017, it was home to Italy’s first Amazon strike. The result was union recognition by management and the stipulation of a plant-level collective agreement on working time and night shifts.
Since then, the unions’ strategy has been to extend this agreement to other areas, especially bonuses, health and safety, and information rights. But the company refused to seriously discuss these questions — and only conceded an annual renewal of the agreement on night shifts in 2019 and 2021. For both Italian and US Amazon management, the agreement following the first strike was only ever meant to be an exception. Moreover, the unions seemed either unwilling or unable to mobilize — nor were fulfillment center staff exactly putting irresistible pressure on them to do so.
One Strike, Different Conditions
Yet the period following this first strike also saw Amazon’s Italian operations grow massively, through a vertical integration process which extended across all European countries.
In 2017, Amazon opened two new Italian fulfillment centers, in Vercelli (between Turin and Milan) and Rieti (near Rome). In 2019, it created another one on the outskirts of Turin and, in 2020, three more in Rovigo (in the northeast), Pomezia, and Colleferro (not far from Rome). It also created autonomous delivery branches, with twenty-five stations where delivery drivers load orders into their vans and start their shifts. This time, unionization didn’t take so long as in Piacenza: the previous strike accelerated organizing efforts, and Rieti and Turin were rapidly unionized. Drivers also organized, first in the country’s wealthiest region, Lombardy — where most of them operate — and then in Rome, Genoa, and Tuscany.
Represented by the logistics branch of the CGIL, CISL, and UIL union confederations, drivers are the newest addition to the Amazon workforce. Until 2015, when Piacenza was the only fulfillment center, distribution had instead been subcontracted to large logistics firms such as UPS and SDA or to the national postal service. But from 2016, the company started building its own mid-range network, with small delivery stations on the outskirts of Milan and then Rome, to receive the boxes prepared and sent from the fulfillment center.
The couple dozen workers operating each delivery station are employed by Amazon, but subcontracted drivers collect and deliver the goods to customers. Today, Amazon logistics subsidiaries employ more than 4,000 workers, plus around 10,000 temps (according to the unions, that is; no official figures are made public). The exact number of drivers subcontracted by Amazon for “last mile” delivery is also hard to judge, given the extreme fragmentation of this workforce (see chart).
Being a driver is even harder than being a fulfillment center worker. Both have to deal with demanding work rhythms, constant control, and a total lack of autonomy. But drivers’ conditions are especially precarious: while Amazon plans and monitors their work (determining their routes, workloads, hours, and evaluations), it doesn’t count them as employees and accepts no responsibility for them.
At the bottom of the chain of command, drivers have to cope with the most intense exploitation and the greatest “flexibility.” Outsourced delivery firms hire them in the thousands during the peak season. These firms are set in competition by Amazon, to deliver more, and faster — a pressure which is then heaped on the drivers.
The productivity bar is raised in each peak season, forcing workers to do ever more to satisfy management demands and digital control. After peak season, the new productivity standard is maintained, but half the workforce is laid off, and the rest have to continue at the higher pace. According to unions, “in the first few weeks of 2019, Amazon drivers were delivering up to twice as many packages per day” as the industry average.
Behind these increased workloads is the fact that Amazon’s market share has grown massively without a commensurate rise in the number of stable jobs. Yet this is also driving complaints among workers, who know their difficult job navigating traffic in Italy’s main cities is undervalued. The intense pressure they are under explains why many of the strikes that first emerged among drivers were spontaneous. In these actions, workers demanded an end to managerial harassment, incessant productivity increases, and the so-called franchise system, which forces workers to pay penalties to their employers for traffic fines or damage to their vans.
The first drivers’ strike in Milan in 2017 was followed by others in 2018. That October, a supply-chain-level collective agreement was signed by the confederal unions and the business association for the outsourced delivery firms — though not by Amazon itself. However, this did not satisfy the workers, and, in January 2019, they organized another walkout, together with the confederal unions, across the Lombardy region. This strike was an important step forward: it involved coordination across this large region, no mean feat among an individualized and fragmented workforce, though it still involved only drivers and not warehouse staff.
State of the Unions
Italy isn’t the only country where Amazon workers have been on strike. Germany was first in 2013, followed by France in 2014, Italy in 2017, and Spain in 2018. These strikes were basically site based. Given the rapid growth of Amazon’s logistics network, unions were unable to keep pace with new openings and unionize all the sites. A partial exception is France, where labor legislation promotes mandatory workplace elections. This encourages a union presence at all sites as well as centralized collective bargaining in the subsidiary that manages the fulfillment centers there, Amazon France Logistique. This favored coordinated mobilization during the pandemic, when unions called for a national strike across fulfillment centers. Yet this did not extend to smaller hubs, or drivers, who are either employed by the separate Amazon France Transport or outsourced.
Across the Alps, Italy’s confederal unions faced the same problem. The firm structure is similar, with one subsidiary in charge of logistics operations (Amazon Italia Logistica) and another for the delivery stations (Amazon Italia Transport), alongside a galaxy of small- and medium-sized outsourced delivery firms. Yet here, unions had more foresight in their strategy for organizing the whole workforce. They realized that last-mile delivery was a weak point of the Amazon network — and the level where exploitation was at its highest. Therefore, the unions, especially the CGIL’s logistics branch, invested in organizing drivers. And, as the first spontaneous walkouts in 2017 showed, drivers didn’t have to wait for the unions’ green light before taking action.
Building a common strategy for all Amazon logistics workers was hardly easy — and the confederal unions’ structure, rigidly divided into industry-level federations, makes coordination among drivers and warehouse staff difficult. For instance, Amazon’s oldest Italian fulfillment center, in Piacenza, is covered by the collective agreement for the commerce sector, whereas the rest of its operations come under the agreement for the logistics and transport sector. This means that warehouse workers, employed by the same subsidiary but operating at different sites, are represented by two different union federations.
Such an artificial division, the product of Amazon’s decision to sign two different collective agreements — and reproduced by the unions’ industry-level structures— not only makes coordination among workers and union delegates harder, but also frames their thinking. Some do not consider such coordination a priority because “workers under the commerce contract don’t have the same problems as their colleagues with the logistics one.”
However, the stakes were high and sooner or later such obstacles needed to be overcome. Amazon is one of the world’s largest employers and is in the vanguard of the digital “revolution.” It is a powerful monopoly that seeks to disrupt the rules of the game, from labor regulations to antitrust and tax laws. The pandemic, and the reorganization of the capitalist economy that followed, have only heightened its power.
For these reasons, this strike has undeniable political importance. But the importance of the unions’ decisive move should be understood within the wider dynamics of Italian industrial relations, especially in the logistics sector.
In the last two decades, Italian confederal unions went through a process of institutionalization — following an embrace of social dialogue and corporatism from the 1990s onward, which fueled a process of incorporation into the Italian state. Unions gained access to policymaking — but in exchange conceded wage devaluation and an end to industrial conflict.
This especially hurt unions’ ability to mobilize the periphery of the labor market — notably, the emerging logistics industry. It also made it harder to resist labor market deregulation, wage cuts, and management’s increasing discretionary power. Workers in the logistics sector, especially migrants, were segregated in a fragmented chain of outsourcing: large logistics companies subcontracted warehouse work to cooperatives under deregulated and often illegal conditions. These cooperatives, especially, blackmail migrant workers — whose permits to live in Italy are tied to their jobs — in order to impose overtime, unpaid working hours, and despotic rules.
The confederal unions took no serious initiative on this front, partly because of the historically good relations between the cooperative movement and the trade unions. Ultimately, this status quo was disrupted by the formation of rank-and-file unions SI COBAS and ADL COBAS. From 2011, these independent unions, made up of grassroots activists and logistics workers, began making waves across Italy, and changed the balance of forces in the sector.
COBAS’s successes owed much to its flexible union structure, which relied on preexisting migrant community ties and allowed migrant workers to take leadership roles. Also key were its weapon of sit-ins and blockades at the warehouse gates. This allowed workers to paralyze the circulation of goods. The pressure considerably improved economic and working conditions and brought a wave of revitalized union organization to the sector. Today, SI COBAS and ADL COBAS are the most representative unions in some of Italy’s most important express and courier firms.
This also brought harsh repression: picket lines attacked by employers’ henchmen or the police, employers’ lockouts, politically motivated layoffs, judicial charges, fines, and trials. Just last week, SI COBAS activists in Piacenza were arrested after a difficult but victorious strike against FedEx-TNT.
Amazon was, in fact, the only firm where COBAS was not able to establish a significant presence. This could be explained by the peculiarity of its employment conditions, the style of HR methods, and, especially in fulfillment centers, the relatively greater promise of stable employment.
This also helps explain the Amazon strike’s political importance for the more established confederal unions. This walkout was triggered by a sudden stop in negotiations, and the unions want to show the multinational their strength. But, loyal to their narrow conception of social dialogue, their aim is to force Amazon back to the negotiating table — to establish the normal industrial relations that exist in other sectors.
Victory is anything but certain — Amazon’s strategy is to avoid any serious engagement, especially when it comes to wage levels and workers’ control over the labor process. Added to that, the Italian branch has no effective decisional power or legitimacy to sign an agreement without the consent of Amazon HQ in Seattle. The company has managed to pursue this strategy successfully so far, and will continue to do so until unions and workers can really hurt the ordinary functioning of logistics operations and stop the flow of goods.
Monday’s strike is historic for the labor movement. But whether it can also prompt a broader revival of the unions depends on workers’ ability to keep the fight going and extend it to other workers, such as temps, Amazon’s call center operators, and workers from other firms. For now, one positive sign was the action by UPS drivers in Milan who refused to deliver Amazon’s products on the day of the strike. Such solidarity is fundamental to overcoming the division among different groups of workers — and building an international front against the multinationals.