Data collection from Roma and migrant communities

To date, knowledge about COVID-19 has largely been shaped by quantitative data.13 This data has reflected biomedical markers of severity and recovery, including caseloads and mortality rates, as well as the outcome of epidemiological models such as the ‘R-number’. Recently, however, scholars have called for more research grounded in critical approaches to communities, as well as in wider frames of well-being.14 This is not just a call for counting the uncounted, but for an understanding of alternative ontological orientations towards emergency healthcare. We situate our empirical findings with the imperative of critical medical anthropology to understand COVID-19 in view of the structural determinants of health, and the deep inscription of social inequalities in the endurance of crisis.15

Our findings draw on research conducted through the ‘Ethnographies of Disengagement’ project.16 This study initially sought to understand reasons for ‘hesitancy’ among Roma, migrant and refugee populations, through engaging with the experiences and perspectives of community members themselves.17 Our focus was derived from a review of the extant, though limited, literature pertaining to COVID-19 vaccines among these groups.18 19 Departing from critical medical anthropology literature, we understood vaccine choices to reflect, among other aspects: populations’ understandings of the state; historical disenfranchisement; regional inequalities and economic marginality. In contrast to dominant frameworks of ‘vaccine hesitancy’, we understood choices to reflect structural barriers to accessing care.

Given that recent policy shifts in the administration of vaccines had led many members of our participant groups to accept a vaccine, we shifted course to understand unfolding experiences of state-led vaccination campaigns. As such, the research charted wider concerns around the experience of vaccination, bureaucracy, individual and communal health concerns. Our focus on the erosion of trust was driven by our interlocutor’s narration of experience.

The study involved a team of four researchers who conducted 84 interviews between October 2021 and January 2022. Research engagements took place with three main groups: Roma communities in Rome, Milan and Catania; undocumented migrants residing in Rome; and people seeking to cross Italy’s Alpine border into France from Oulx, province of Turin. While these groups are diverse, they have all been included in recent policy declarations to construct inclusive approaches to vaccination.20

In Rome, participants were largely young men from Albania, Peru, Ecuador, Morocco, Bangladesh, Venezuela and Indonesia. Participants lived in housing ‘occupations’ or rented apartments with other co-nationals, and had done so for several years, earning a living through informal labour including as caregivers, factory labourers, restaurant workers and cleaners. In Oulx, interlocutors were single male travellers, or family groups, originating predominantly from Afghanistan, Iraq and countries in Central and West Africa. In Rome and Oulx, all lacked formal Italian citizenship. Roma communities in Milan and Rome were from the ex-Yugoslav republics and Roma people in Catania were originally from Bulgaria and Romania Most of the Roma interlocutors, who were both male and female, were employed in informal labour as street vendors, drivers, domestic workers and caregivers. Many Roma were Italian citizens although others were waiting for legal status.

Bearing in mind the heterogeneity, mobility and access restrictions surrounding the communities included within this research (in the content of COVID-19 restrictions), approaches to addressing vaccine participation required varying outreach strategies and methodologies. In Rome, the ethnography of undocumented migrants required the collaboration of health authorities, Non Governmental Organisations (NGOs) and self-organised groups, and involved 20 open-ended interviews. Our methods with ‘people on the move’ adapted to a context of hyper-mobility where migrants often spent less than a day in refuges while preparing to cross the border into France, 22 open-ended interviews were conducted, the timing and format adapted to the rhythm of migrant schedules. The case study among Roma communities consisted of in-person participant observation and 42 in-depth interviews (16 in Milan, 8 in Rome and 18 in Catania). Two additional workshops were conducted to verify and co-produce our findings with Roma communities and leaders.

An additional 21 semistructured interviews were conducted with trusted representatives from respective communities, including faith and community leaders, social workers, doctors, health workers, scholars, policymakers and volunteers involved in vaccination campaigns between December 2021 and February 2022. This wider network of participants provided crucial context as to the changing dynamics of vaccination campaigns, as well as insights into communal resistance over the course of the pandemic.

Undocumented migrants in Rome: vaccine uptake tied to citizenship

At present, between 500 000 and 600 000 undocumented migrants live in Italy.21 Little is known about these groups’ attitudes towards the vaccine.5 Living without citizenship or formal claims to state welfare, these migrants have been particularly difficult to access: existing data does indicate, however, that the clinical and socioeconomic burdens of the pandemic have been greater among migrant groups with higher rates of pre-existing health conditions, and without economic safety nets.22

Across Italy, undocumented individuals face structural barriers to accessing routine public health services.23 Pre-existing impediments had a significant impact on migrants’ interpretation of, and willingness to engage with, the vaccination campaign. Yet, additional barriers arose through the construction of a campaign that relied on citizenship to access vaccines. Between January and July 2021, non-Italian citizens (lacking a National Insurance Number) could not register for a vaccine.24 By September 2021, it had become possible to register for a vaccine using an Straniero temporaneamente presente (STP) code, a code reserved for visitors and non-nationals. Yet, early exclusion, limited information and regional inequalities still created uncertainties among migrants about accessing a vaccine.

Further obstructing access was the tying of vaccines to the Green Pass certification, which in Italy granted the right to work and to access particular transport routes. Across sectors with heavy migrant labour, including domestic or cleaning work and manufacturing, employers actively enforced the Green Pass as a prerequisite to work. These requirements were often used by health workers to incentivise vaccination. For example, one health worker in Rome explained: ‘We used to use the Green Pass as an incentive to vaccination. […] To convince them (undocumented individuals) we used to say ‘without it you won’t be able to work’. That’s how we would convince them, and a lot of people got vaccinated for that reason, not for anything else.’

In this sense, Green Pass certifications were viewed not as an instrument not of public safety but of state coercion. Among our respondents, attitudes towards COVID-19 vaccination were frequently characterised by frustration, anxiety and a widespread feeling of having been forced to obtain one. These feelings were amplified by reported delays in obtaining certification following vaccination. As a consequence, many vaccinated workers had to pay for COVID-19 tests (which cost a minimum of €15 at local pharmacies) several times a week while waiting for their Green Pass; when unable or unwilling to do so, many lost their jobs. This account reported by one respondent is illustrative: ‘You can’t imagine how angry the Kurdish people were. We went to get vaccines and then the Green Passes didn’t arrive. They got angry with us and said ‘what happened to our certifications? We need to work!’’.

These findings highlight how state approaches to tackling the pandemic have generated both the proliferation of bureaucracy and subsequent policy contradictions posing as barriers to vaccine participation and social justice. Though many migrants had received vaccines, their individual autonomy was compromised in the experience, undermining their trust in the state.

People on the move in Turin: vaccine uptake tied to mobility

For the second group of migrants—recently-arrived individuals, including asylum seekers and refugees—Italy served as a transit country for journeys intended to end in other European countries. Vaccine passes presented an obstacle to onward movements. Accordingly, decisions to receive a vaccine were described in pragmatic terms, driven primarily by mobility needs, rather than fears of contracting COVID-19. When asked why they had decided to obtain vaccination, the vast majority of our respondents reported doing so in order to facilitate their journey, or to avoid problems with police. In Italy, Green Passes were a prerequisite to accessing trains across the country, and thus a necessary prerequisite for internal movement, as well as international border crossing. While the border between Italy and France has historically been relatively porous, over the past 7 years it has become one of the most policed in Europe. This meant encounters with authorities were likely, and vaccine passes were necessary.

Reflecting the diversity of migration routes, many of our respondents had accepted a vaccine along the way. For example, one young Afghan man explained his motivation: ‘For me, I got the vaccine in Greece, because I thought that if the police ask for a vaccine certificate, at least I will have no problem.’ At the time of our research, many people undertaking the crossing were fleeing from conflicts in Afghanistan and West African countries, and had received vaccines in Turkey or Bosnia, or at ‘camp’ stops along Central and Eastern Mediterranean routes where humanitarian organisations were distributing vaccines. In many cases, people had received a Sinovac or Sputnik vaccine, which were not recognised by the Italian state as offering immunity (or making one eligible for a Green Pass). Issues for these people were thus not of vaccine hesitancy but of vaccine recognition.

Our respondents often conflated health workers with state officials and border police. All such agents of authority were considered obstacles to onward movement, as most migrants were seeking to pass through Italy undetected and were well aware that, had they tested positive for COVID-19, they would have been made to quarantine, thus preventing their journey. Such concerns were prominent among male migrants, travelling alone or leading their families, since there was immense pressure to complete journeys.

Bearing in mind that many of our interlocutors were fleeing violence, and embarking on long, expensive and treacherous journeys, COVID-19 often lay at the margins of larger attempts to find safety. As one young man who had fled from Afghanistan summarised: ‘Only the Taliban are a big problem in Afghanistan.’ Even so, stringent regulation among border authorities meant that for these people vaccines were primarily understood in relation to state governance, as a manifestation of attempts by different European countries to restrict movement. Again, frequent feelings of coercion were reported as people perceived they had been forced to obtain a vaccine by law enforcement. In the words of one Senegalese man: ‘I got it because the police told me to. I have no power; the government has power.’

Roma communities: vaccine uptake tied to structural discrimination, past and present exclusion

For Roma communities, recognised as Europe’s largest ethnic minority of approximately 12 million people, the COVID-19 pandemic was particularly challenging.17 Our research found that three main and interrelated factors—structural discrimination, negative past and present encounters with the state and mistrust in public authorities—have influenced Roma attitudes towards COVID-19 vaccination. Yet, no official data exist on Roma vaccine uptake, or rates of infections, hospitalisations or deaths.17 Past experiences of structural discrimination, alongside present hardships brought by the pandemic, greatly influenced Roma responses to COVID-19 vaccination.

Known by the derogatory term ‘zingari’, some Roma in Italy live in ‘nomad camps’ (campi nomadi) also called ‘villages’, usually located on the outskirts of cities created as a result of local and regional government regulations in the 1980s and 1990s.25 26 The alleged illegality of these camps, created as temporary housing solutions for Roma perceived as migrants and people on the move, has long caused frequent police visits, reinforcing the social and physical boundaries between camp dwellers and the rest of the population.27 Seen as spaces of contagion, the securitisation of Roma camps increased even further with the onset of the COVID-19 pandemic.

Due to a perceived reluctance to adhere to pandemic measures, Roma have been portrayed in public discourse as a threat to public health, which has reinforced prejudice against them.28 During lockdown military presence in Roma camps increased as an intention to control the spread of the virus, while evictions of camp dwellers continued throughout the pandemic.29 30 As in other European countries, lockdowns for Roma communities took place on a settlement and community level rather than on an individual and household basis.30 Such policy responses diminished Roma trust in government initiatives further. Research participants articulated mistrust in government and health authorities due to historical and present discrimination which merged to form a resistance against vaccination when it became available. One research participant shared: ‘They (the government) offer me the vaccine while taking away the only home I have known since I was a child.’

Facing poverty, unemployment, precarious work in the informal economy, lack of adequate housing and accommodation and impending evictions intermeshed with the detrimental effects of the pandemic.17 29 31 Due to their status as informal workers, some of our Roma research participants were not covered by social welfare or health insurance. The precariousness of informal work led them to rely on communal solidarity and mutual support, which was often limited and worsened their socioeconomic situation. The social restrictions and isolation disrupted vital communal networks of mutual support within and between Roma households. Many of our Roma interlocutors in both formal and informal settlements found themselves struggling to access food, medication and general services. In the words of one interlocutor: ‘I used all my savings, everything, for my family and friends. We could not go to the market to sell and earn money to feed ourselves.’

Structural barriers also contributed to mistrust in the vaccination campaign. Mostly living on the outskirts of cities, research respondents relied on public transport to access services and employment. The limited access to vaccination centres and the lack of state socioeconomic support represented barriers to vaccination, even if research participants wanted to be vaccinated. Found in a precarious situation, the Roma in our research often prioritised vaccination as the means to access employment and to ensure socioeconomic survival. Many considered the risk of COVID-19 in comparison to the prominent need to fight economic precarity summarised as follows: ‘I either die of COVID-19 or of hunger’. Moreover, subsequent vaccine mandates requiring personal information aggravated prior fears of eviction and deepened mistrust of government authorities. The lack of residence and health records also meant that access to public health services and to vaccination were limited, which led our interlocutors to seek costly private healthcare provision where vaccination was not available.

Crucially, the lack of Roma representatives in the vaccination campaign also contributed to mistrust towards health professionals and government initiatives. An elder in Milan pointed to the vaccination campaign seen by his community ‘as a solely non-Roma initiative’. Despite this, Roma who had access to and an established relationship with a general practitioner, were more likely to be vaccinated and to trust health professionals. Instances in which there was communal involvement, including a Roma health mediator, elders, community activists and religious leaders, resulted in increased vaccine uptake.

Vaccine acceptance and the production of mistrust

While to date mistrust has been largely understood in relation to unvaccinated persons, our research shows that mistrust belongs to the vaccinated as well. In this way, our findings extend studies which link vaccine regulation to mistrust among unvaccinated in European countries.32 Among these diverse communities, many migrants and Roma people perceived measures to regulate immunity as coercive, interpreting policies as a continuation of everyday experiences of discrimination or neglect that characterised everyday life. That (continually changing) bureaucratic shifts were ill-communicated to minority groups, that systems themselves were fraught with delays, and that health-workers tying vaccination to the right to work as a form of persuasion, all served to compound feelings of frustration among our interlocutors.

Across our respondent groups, the vaccine was linked to frustration that individual autonomy had been compromised in a new realm of life. Current literature has assumed that minoritised groups mistrust the state as a rule, and so resist vaccination.33–36 We found, however, that disempowerment resulted in the suppression of individual choices regarding vaccines: many at the margins felt compelled to comply with vaccine mandates, lest the state further constrain their economic activities. Indeed, such disenfranchisement was widely expressed in sentiments which indicated that individual resistance was secondary to performing compliance. Additionally, many of our interlocutors stated that the lack the energy or resources to comply with additional testing requirements, in lieu of obtaining a vaccine pass.

Similarly, we found that despite accepting vaccines, fears persisted both at the individual and community level about the biological effects of the vaccine. While current literature reports the fear of side-effects as a barrier to vaccination, we found that individuals continued to fear long-term effects following vaccination.37 38 Many interlocutors feared the side-effects of the vaccine and potential impacts on long-term health conditions due to high levels of pre-existing illness. For many who had accepted the vaccine, fears persisted beyond the manifestation of physical side-effects. Among Roma populations, collective narratives—generational in origin—that linked the vaccine to state-led extermination also continued after communities displayed healthy responses to the vaccine. Collective fears were related to historical and contemporary discriminative practices and structural inequalities.

Many our interlocutors explained that accepting a vaccine had not brought any change to the position of economic and political marginality that they experienced in everyday life. Some lamentations need to be understood with reference to the concurrent policy environment. COVID-19 vaccines were delivered amid highly contradictory state policies: while protection from a virus was promoted, other forms of well-being were compromised. For example, migrants expressed frustration that informal employers instigated checks on Green Passes but failed to alter poor working conditions (which had produced greater vulnerability, and higher death rates during the pandemic). In one long-established Roma village, eviction orders were served by the police at the same time as vaccines were rolled out by public health authorities in vaccination centres. In essence, vaccines arrived at a complex nexus whereby wider socioeconomic determinants remained unchanged, and in some cases, structural marginalisation was only deepened. For many—even those who did not question viral protection—a vaccine did little to alter prospects for well-being.

Ultimately, those at the bottom of the socioeconomic ladder (be they vaccinated or not) have felt the repercussions of mass restrictions and sanctioning during COVID-19, the most. Our research revealed stories of loss, unemployment, hunger, and as a result, mistrust. This cycle created a landscape whereby vaccine policies were liable to be viewed with scepticism, while the pandemic continued to exacerbate wider challenges to health and well-being. Even though many respondents had accepted a vaccine, this did little to dislodge their mistrust in state-led health campaigns. As of this writing, such broken trust is unlikely to be rebuilt in a swift, lasting, or robust way.

Principles of engagement

Considering the longer temporalities of mistrust that can accompany vaccination campaigns, we propose that principles of equitable engagement that prioritise longer-term trust-building and diversity. We depart from the cautions proposed by Gamlin et al that, ‘[i]f the pandemic is not homogeneous, our responses to it cannot be either.’15

While scholars have recognised the need for diversity, this has largely been translated into interventions which adapt health messaging to various linguistic or sociocultural contexts.34–36 39–42 Yet, our evidence suggests that mistrust in vaccines is not only driven by misinformation, but by wider processes of exclusion and discrimination. Thus, we propose a radically different starting point for policymaking that privileges diversity rather than uniformity and prioritises longer engagement timelines than simply registering jabs in arms. To this end, four key principles emerge.

First, it is important that vaccine strategies are adapted to the social topographies of minoritised groups. This invariably requires tapping into mediators of social and solidarity infrastructures, to encourage community ‘buy-in’ of mass vaccination and health improvement.43 Tapping into networks of kinship and care though trusted local partners—including community representatives and champions, voluntary sector workers, and religious organisations—are key means of encouraging vaccination uptake. Our research found that trust rested within family and communal relationships, and that these networks are key contact points to influence understandings about vaccines and health protection.

Second, it must not simply be assumed that extant infrastructures can be harnessed to deliver health information. Many support organisations rely on volunteers, who may already be at capacity following the demands of the pandemic. Safehouses (refugios) assisting migrants, for example, already rely on strained funding, and struggle to equip migrants for Alpine journeys in short windows of time. More generally, community organisations have suffered from funding cuts following not just the economic shocks of the pandemic but from decades of austerity as well. It is thus important to invest in these networks, and fund additional community mediators, as well as doctors and nurses, to take on health engagement activities.

Third, it is essential that approaches do not adopt a directive approach but instead create space for deliberative engagement. It is essential that external actors allow community leaders, as well as their members, to openly voice fears about the risks and side-effects of vaccination, as well as longer historical experiences of medical racism and negative encounters with the state/healthcare system.44 Communities are not homogenous, but are in themselves structured by power dynamics. Our research suggests that it is also important to address gendered inequities found in fears of side-effects, which were particularly prominent around women. It is important to include assistance or signposting to relevant onward support such as legal assistance with asylum, council support for housing, relevant channels to contest employments and psychosocial support.

Fourth, it is important that campaigns provide a space to address harms perpetuated by the pandemic. Both as individuals and as collectives, our respondents faced issues of priorities in direct competition with one another, such as accessing employment, medication, food, documentation and safe passage. Hence, reconsideration of the ‘mass’—at times even arbitrary—nature of quarantine is crucial, and such measures must account for the needs of the affected populations such as access to amenities and services, aid, healthcare, education and employment. Rather than conceiving of COVID-19 as a ‘great equaliser’, it is critical to acknowledge inequalities the pandemic has perpetuated. Understanding the entanglements of pre-existing inequalities with the new forms of marginalisation the pandemic has induced is a key starting point for engagement.