Evangelical Christianity as Infrastructure in Brazil's Penal System

Abstract This article contends that Brazil's evangelical Christian networks increasingly function as penal infrastructure. Since the 1990s, the scale and scope of evangelical involvement in the criminal justice system have grown significantly. One clear result is that the capillary relationships that constitute Christian community now mobilise resources to support or even substitute the basic functions of punishment. I draw on fieldwork in Rio de Janeiro to understand this shift and its broader implications for the Brazilian project of incarceration. I also make a general claim for thinking with and through infrastructure as a pathway to understand penal governance.


Introduction
The presence of Christianity in Rio de Janeiro's Patronato Magarinos Torres 1 hits you from the moment you walk through the front door because the reception area is furnished with church pews. This is the institution responsible for overseeing all punishments in the state of Rio that are classified as 'restrictive of liberties'. While its role overlaps with that of a parole office, and parolees make up the majority of its clients, it also oversees sentences involving house arrest, community service, probation and weekend detention. The pews arrived a few years before I began fieldwork. They were a donation from the Igreja Universal do Reino de Deus (Universal Church of the Kingdom of God) after the new director of the Patronato, Aline, 2 reached out to them for help in making some much-needed reforms. By the time I began my research, clients would spend most of their visit in these pews as they waited, often for hours, to be seen.
The donation was a small thing. But the fact that it took place at all was the result of a set of transformations that have shaped the relationship between evangelical Christianity and punishment in Brazil. When Aline reached out to her contacts from the nation's largest Neo-Pentecostal denomination for help, she took part in an emerging form of prison governance that has relied on collaboration with evangelical communities to make up for administrative and state neglect. These pews served a rather innocuous function but, in being repurposed, they also reconfigured the Patronato, influencing how both staff and clients moved through a space that now resembled a small church. It is this dynamic between the repurposing of evangelical networks, on the one hand, and the reconfiguration of punishment in Brazil, on the other, that I address here through the notion of evangelical Christianity as infrastructure.
This article claims that the support networks that operate within the evangelical community have now been called upon to supplement or substitute some of the basic functions of Brazil's penal system. The state fails to replace chairs; to provide electricity or water; to maintain the structural integrity of buildings; to attend to its basic legal obligations; and to guarantee the survival of the incarcerated population. These are problems of its own making, the product of a massive project of carceral expansion and the deliberate corrosion of virtually all aspects of punishment that fall outside the narrowly defined ambit of 'security'. Increasingly, staff and individual administrators within the criminal justice system turn to missionaries, churches or religious non-governmental organisations (NGOs) to patch up these frayed ends. Together, these actors produce new channels for the provision of a whole range of resources, from water and clothing to construction materials and legal counsel. Initiatives that aim to mitigate the ongoing breakdown of the physical and bureaucratic structures of punishment ultimately become integral to those same structures. Thus, missionaries and other representatives of the faith come to work in concert with walls and pipes, departmental budgets and filing systems. The aggregate result is that evangelical Christian networks become essential for the basic operations of punishment and for the reproduction of penal governance.
Similar stories repeat themselves across Latin America. Despite an ongoing boom in prison construction, the physical structures of confinement buckle under the weight of a security politics in the region that has, in the last 30 years, expanded the scope of policing and produced an unprecedented zeal for imprisonment. 3 In Brazil, this security regime has also emerged within a landscape of infrastructural neglect and disrepair that extends far beyond prison walls. This was particularly the case in Rio de Janeiro during my fieldwork. In the wake of the 2014 FIFA World Cup and 2016 Olympic Games, the state governor declared a 'financial disaster', leaving half-finished construction projects and a concrete environment that was literally crumbling around residents.
But for those living in Rio's urban peripheries, the absence or dysfunction of basic infrastructure is nothing new. In fact, these communities are often defined by such an absence. In Rio, as with many other Brazilian cities, residents of these peripheries are largely and disproportionately Black, the result of decades of residential segregation, forced evictions and rural-to-urban migration. As Jaime Alves reminds us, these extensively policed neighbourhoods are, like prisons, sites of Black confinement. 4 What's more, the visceral effects of infrastructure's limited reach (pollution, odours and bodily pain) are often understood by Brazilian politicians or its largely white upper and middle classes as inherent properties of residents themselves. 5 This shift in responsibility pathologises poor, Black communities and their residents as 'problems' to be overcome. Understanding evangelical Christianity as infrastructure requires that we situate it within this broader landscape of Brazil's 'penal democracy'. 6 Yet it also means recognising that this is not a landscape of abandonment; instead, as I demonstrate below, breakdowns and ruins are also sites for reworking what constitutes infrastructure in the first place.
My argument draws on the ongoing 'infrastructural turn' within the social sciences of the last two decades. 7 In particular, the claim within this literature that infrastructures are necessarily socio-technical assemblagesthat they integrate human action and politics within the operation of wires, pipes and asphalt, and vice versa 8presents an opportunity to rethink who or what makes things move within Brazil's carceral project. As we will see below, an ad hoc fundraising network built through church contacts can, at times, function like a series of pipes, since both have the potential to act as vessels for water. To approach evangelical Christianity as penal infrastructure means recognising that the ties evangelicals cultivate with one another, and with institutions of confinement, are also conduits for the distribution of goods and services.
Infrastructure invites us to view these ties along a horizontal plane, as if on a map or schematic diagram. This perspective reveals the work of reticulation, the integrating 9 or splintering 10 of actors by a broader circuitry within which goods move. But infrastructure is also a concept of verticality, one that draws our attention to the undersides of social and political life. It therefore presents an alternate point of entry into analysing penal governance, a central concern in studies of punishment within Latin America. Accounts of 'self-rule' 11 or 'co-governance' 12 in the region's prisons are currently central to how scholars understand these institutionsso much so that the negotiation of power between incarcerated people and correctional officers has arguably become the single defining characteristic of a Latin American 'model' of imprisonment. 13 While the terms vary, 14 this line of inquiry stresses that the social order of prisons, and the forms of control that maintain them, are constantly negotiated between staff, administrators and the imprisoned. Such negotiations emerge from an extremely small number of guards relative to the prison population; 15 the necessity of self-organisation to meet those basic needs of survival that the state withholds; 16 and, in many cases, the consolidated political power of networks that tether survival to organised crime, networks that are often glossed in English as 'gangs '. 17 While this literature offers many important insights, it also risks flattening our understanding of governance by reducing it to the dynamics of survival and control. In particular, the scope of this analysis is often limited to a dyadic relationship between correctional officers and the imprisoned. One notable effect of this is that 'governance' is implicitly gendered as masculine, not simply because these studies overwhelmingly focus on incarcerated cisgender men, but also because the work of womenincluding missionaries, visiting family members and the feminine sphere of 'treatment' staff like social workersis assumed as peripheral to the creation and maintenance of order. 18 Brazilian scholars have offered important correctives to this literature, pushing back against an exaggerated notion of incarcerated people's agency 19 and calling into question the extent to which their forms of relationality might constitute a 'structure'. 20 My shift to infrastructure here is intended as a complement to these critiques. I want to refocus our attention towards the actors and relationships that subtend governance, and thereby re-evaluate how its basic conditions are maintained and reproduced.
To return to the pews: Aline's recourse to the Universal Church was only possible thanks to a consolidated web of relationships that she and others had cultivated between churches, NGOs and staff like herself. These relationships were largely (although far from exclusively) produced and maintained by women. The 13 David Skarbek, The Puzzle of Prison Order: Why Life behind Bars Varies around the World (New York: Oxford University Press, 2020). 14 For instance, although both anglophone and lusophone scholarship engages with these questions, the latter rarely uses the term 'governance' (governança) and is more likely to use the frame of 'order' (ordem  19 web, in turn, is the product of a decades-long intensification of evangelical participation in the criminal justice system. Brazil's Lei de Execução Penal ('Law of Penal Execution'), legislated in 1984, disturbed the Catholic Church's de facto monopoly on religious assistance within the nation's prisons. It guaranteed and protected the religious freedoms of both the imprisoned and the missionaries who visited them. Soon after, 'traditional' evangelical churches like the Baptists began visiting prisons around the nation; they were followed by  In this article I use the term 'evangelical' to refer to the representatives of various congregations who are involved directly with the prison system. This includes pastors and missionaries, but also workers and volunteers from religious NGOsas well as some prison staff, like Aline, who speak of their work and faith as bound to one another. Today, Pentecostal denominations represent the largest and most visible Christian presence within Rio's prisons, part of a spiritual movement that Andrew Johnson has identified as 'prison Pentecostalism'. 22 My use of evangelical, by contrast, parallels evangélico as the preferred term for both the individual identities and the broader spiritual movement in which missionaries, staff and incarcerated people alike took part. Evangelical spirituality steadfastly refuses, at least in principle, any moral distinction between those who are and those who are not incarcerated. Nevertheless, for the sake of clarity this article generally uses 'evangelical' to refer to the latter.
The role that these churches play is not entirely new. In many cases, evangelicals have assumed a position formerly occupied by Catholics in the previous two centuries. Catholicism has not disappeared; arguably the most prominent example of Christian investment in punishment comes from Brazil's Associação de Proteção e Assistência aos Condenados (Association of Protection and Assistance for Convicts, APAC) model, where prison administration is given over to a Catholic-founded religious NGO. 23 But despite the international attention given to APAC prisons, their reach is limitedthere are no such prisons in Rio, nor in the majority of Brazil's other states. And while the Catholic Pastoral Carcerária (Prison Ministry) plays a central role within the national campaign for decarceration, its presence and influence at the ground level has waned. 24 The processes that I describe here are ongoing and uneven. Their immediate effects are often difficult to measure, since evangelicals most often play a stabilising role, one aimed at mitigating or avoiding breakdowns. Actors situated across  The Pastoral Carcerária does, however, play a critical role in facilitating researcher access to prison systems, most notably in São Paulo. See Rafael Godoi, Marcelo da Silveira Campos, Fábio Mallart and Ricardo Campello, 'Epistemopolíticas do dispositivo carcerário paulista: Refletindo sobre experiências de pesquisa-intervenção junto à Pastoral Carcerária', Revista de Estudos Empíricos em Direito, 7: 1 (2020), pp. 143-58.
Brazil's carceral project struggle with the increased presence and scope of evangelical involvement in the penal system. This includes evangelicals themselves, who often critique their own positions and express frustration in being shouldered with what they see as responsibilities of the state. As such, this is not an account of religious groups hijacking institutions of punishment. Evangelicals' influence is often circumscribed precisely because their attention and their resources are re-routed towards crisis management and the maintenance of basic services. But they are nevertheless investedsometimes willingly, sometimes notin both the imprisoned population and the institutions that hold them. Where many of these investments are aimed at supporting the well-being of the imprisoned, they also allow the state to experiment with new ways of maintaining its own authority, even within a system that often seems close to breakdown.
The evangelisation of prisons in Brazil, and in Latin America more broadly, has long been recognised by scholarship. Researchers have emphasised the labour that incarcerated people and missionaries invest in the production of new subjectivities, particularly through conversion. 25 Another flank of inquiry focusses on forms of self-segregation, self-governance and codes of conduct that regulate incarcerated communities of the faithful. 26 While earlier work was often concerned with whether evangelical practice was 'merely' performative, 27 recent studies have developed more granular appraisals 28 and critiques 29 of the politics of, and entanglements between, Christianity and punitive violence.
Here I move the conversation towards issues of distributive politics rather than those of theology, even as I recognise that the two are inseparable. I do so by drawing on 24 months' ethnographic fieldwork conducted in and around the prison system of Rio de Janeiro between 2014 and 2018. As part of this project, I visited the Patronato regularly during the course of its working day, and often stayed after it had closed for clients. In addition, I conducted interviews with staff, volunteers and clients. The Patronato is something of an exception within the broader constellation of units that make up Rio's penal system. It is not a prison, and with no direct custodial function its rhythms of work, as well as the particular challenges that staff face, are unique. But as I will demonstrate here, the institution offers one face of a much broader infrastructural politics that affects the entire system. I make these connections clear by drawing on my work within a 'closed' men's prison that I call René Dotti. By charting this path, my aim is to demonstrate that evangelical Christianity as infrastructure represents an ongoing experiment in penal governance that both shores up the project of incarceration and renovates its underside.

Mapping the Flows
The Guandu River, born from the rains that fall along the Serra do Mar, supplies the vast majority of potable water for the city of Rio de Janeiro. But in early 2015, during the worst drought since record-keeping began, this supply was drying up. The response by prison wardens across the state was to tighten their rationing of the water supply. The practice was not entirely new; for decades, officials had restricted the flow to certain times of the day. In most units, imprisoned people would take this opportunity to fill plastic gallon containers for future use, a practice which carried the risk of contamination. 30 The Companhia Estadual de Águas e Esgotos (State Water and Sewerage Company, CEDAE) continued to deny the longstanding accusations that it restricted water flow to prisons or that it gave priority to the surrounding residential areas. Officials from the prison administration insisted that they were receiving less water. Regardless of who was to blame, supply was now down to one large bucket per person per day in some units, to be used for drinking, washing and bathing.
For Josielle, this moment marked a turning point. As a Baptist missionary who volunteered in a women's prison, she felt that the warden, who was not a 'believer' (crente), treated the evangelical churches that visited the unit as clients rather than 'partners' (parceiras) within the system. But given the dire state of the water supply, and with no forecasts for rain in sight, the warden asked Josielle and other missionaries for help. Collectively, these volunteers pooled money from their respective churches and solicited donations from congregation members. They used these funds to deliver potable water by truck into the prison's water tower. A few months later, as she recounted this story to me, Josielle was still working through her ambivalence about the event. On the one hand, she was pleased that, by addressing an emergency, she and the other missionaries had built a new level of trust with the warden. Now, she explained, the communication between them was far more open and respectful. On the other hand, she also felt that the warden had exploited her goodwill. 'This isn't how a prison is supposed to run', she reasoned, irritated that her time and resources had been diverted from spiritual work with those incarcerated in the prison. And she worried that more of these problems would arise, problems that she understood as the responsibility of the prison administration, but that she knew would continue to be passed onto her.
Both missionaries and staff within the penal system use the language of parcerias ('partnerships') to describe their relationships with one another. At least ideally, parcerias would suggest mutual support, respect and collaboration in the pursuit of shared goals. The word buttressed a deliberate strategy of evangelicals to avoid open conflict with the prison administration. 31 Josielle had learned, though, that it could also produce new forms of dependency or exploitation. As we will see, the form of these calls for help repeated itself over the entire prison system. As they multiplied and morphed into responses to more chronic failures, the line between emergency actions and the baseline operations of the penal system was becoming increasingly blurred.
Since Brazil's most recent transition to democracy in the 1980s, the nation has embarked on a massive project of carceral expansion. The number of the imprisoned rose from an estimated 90,000 in 1990 to over 780,000 in 2019a growth of over 700 per cent. 32 This growth has targeted young, Black Brazilians, particularly those living in urban peripheries; while the vast majority are held in men's prisons, both the absolute number and proportion of incarcerated women have rapidly increased. 33 Although this rise is greater than that of Brazil's neighbours, it exemplifies a broader shift over the last three decades in which Latin America has moved from some of the lowest rates of imprisonment in the world to its highest. 34 In Rio de Janeiro, this transformation has strained virtually all aspects of the penal system, from basic upkeep to the work of administrators, correctional officers and 'technicians' (an umbrella term referring to social workers, psychologists and psychiatrists). But while breakdowns proliferated across many prisons, they took a specific form in the Patronato. If we approach imprisonment as a logistical 'flow' of bodies across a series of institutions, the Patronato occupies the end point of this process, located on the threshold between punishment and the end of a criminal sentence. Since the number of clients (parolees or others serving noncustodial sentences) depends on the rate at which people are released from prisons, shifts further up the conveyor belt of punishment can have huge impacts on the rhythm of work here. In 2016, for instance, a few months after the leader of Rio's Court of Penal Administration began advocating for judges to ramp up parole authorisations, the institution's workload effectively doubled -Aline estimated that her staff were now collectively attending approximately 10,000 clients.
Staff at the Patronato would constantly reconfigure their work to respond to these fluctuations. This was particularly notable with the technicians, who comprised a minority of staff even though they carried a large share of the institution's administrative responsibilities. The social worker Katia, for instance, explained to me during an interview that over her six years' experience in the office 'every instant -I wouldn't even say every month, every year, but every instant -I have to adapt, find a new way to work'. The central duty of technicians was to provide new clients with an 'orientation', a discussion that lasted between five minutes and half an hour and that explained the rights and restrictions of a particular sentence while also quickly assessing the clients' needs. These orientations used to be one-on-one. But when I began fieldwork in 2014, staff were calling in two or three people at a time. By 2017, they were giving orientations to up to eight clients simultaneously. To protect the privacy of clients, they now avoided overly sensitive questions; instead, they ran through a standard script and asked for a show of hands to identify who was in need of work, documentation or shelter. 32 Infopen, Levantamento nacional de informações penitenciárias (Brasília: Ministério da Justiça, 2020). Many staff, and almost all of the techicians, had decades of experience within the penal system. 35 From this vantage point, they spoke of their nostalgia for a previous time, one in which they maintained some semblance of a personal or therapeutic relationship with clients. This discourse often struck me as overly romanticwhile conditions had clearly deteriorated, there was no time in living memory when Rio's prison system had not been overcrowded, for instance. But the narrative helped reinforce the identity and mission of technicians and other staff whose vocations and professional self-image were wrapped around an ethics of accompaniment.
Within the United States, parole officers often negotiate between competing roles of 'cop ' and 'counsellor'. 36 In the Patronato, by contrast, where nobody had any direct punitive control, we might identify the fundamental tension as between counsellor and bureaucrat. 37 Documents move through different channels from water, but both are necessary for the daily governance of punishment. To ensure this movement, staff would outsource their other responsibilities to churches and church-based services. Regardless of their own faith, many understood that the closest available analogue to the kind of accompaniment they understood as necessary for those leaving prison was that offered by evangelical communities. What both groups shared was a sense of working for the incarcerated population, rather than the prison system as a whole. As two domains dominated by women, technicians and missionaries also shared in a language of maternal care, where motherhood offered an explanation of both their motivations and their impact on incarcerated 'children'. So for those looking for work or a place to stay, Patronato staff would suggest a particular group located near the client's neighbourhood. This routinised outsourcing allowed the Patronato to at least partially maintain a semblance of fulfilling its legal responsibilities for providing social assistance. But as it sought out new parcerias and expanded the scope of existing ones, the nature and politics of the institution shifted in more profound ways.

Making Partners, Making Infrastructure
The decentralised and extensive nature of evangelical community life makes it difficult to map out. Yet these two qualities also help to explain how and why it has been brought into the fold of penal governance. With some notable exceptions, most evangelical denominations operate as loose associations between autonomous churches, each under the leadership of one or more pastors. 38 This apparent fragmentation nevertheless supports a thick ecosystem of collaboration and coalition- 35 The last round of permanent employment for technicians occurred in 1998. Since that date, the administration filled vacancies through short-term contracts. During the 'crisis' of 2016-18, these contracts were not renewed, so most of the technicians who remained had at least 20 years' experience. While correctional officers worked in the Patronato, they had no power to punish or detain; their custodial responsibilities were limited to informing the courts when a client did not show up on their due date. 38 The Universal Church of the Kingdom of God, Brazil's largest Neo-Pentecostal denomination, is an important exception, with its vertically structured hierarchy within and between churches. While it had a large presence within the penal system, the church was seen by other denominations, and by many prison staff, as insular.
building. NGOs link churches across denominational lines and draw on their resources for short-and long-term initiatives. Job opportunities travel within and across congregations by word of mouth. Those in need are referred to evangelical health providers, counsellors or social workers, who offer their services for free or at a heavy discount. Pastors collaborate with lay evangelicals working in small businesses or larger corporations to secure donations. Others secure financing from overseas churches, particularly those in the United States; or by participating in Brazil's Christian media industries. Progressive evangelical NGOs also apply for grants from international philanthropic organisations. But the bedrock of this economy is the dízimo or tithe, the obligation of laypeople to donate a portion of their income to the church. The combined result of these processes does not match the scale and intensity of state institutions. But it does reach marginalised territories and populations, and often proves more efficient, reliable and flexible in providing services while presenting minimal barriers to access.
There is no universal desire within the evangelical community to mobilise these networks in service of either the prison system or those impacted by it. The same punitive streak that runs through Brazil as a whole is also present here, including a violent hostility towards those labelled as criminal, as well as support for a spectrum of punishment that ranges from mass incarceration to extrajudicial killings. 39 Having said this, evangelicals in Rio de Janeiro are far less likely than others to support the Brazilian dictum that 'a good criminal is a dead criminal'. 40 And the greatest civic interest in, and engagement with, the criminal justice system comes from these congregations. Pastors, missionaries and volunteers alike ground their engagement with the penal system in the recognition of needthat is, the social and spiritual needs of an incarcerated population who would otherwise be cut off from the word of God and whose daily life they see as inflected with loneliness and guilt. They were also motivated by the imperative of salvation, a central responsibility of all practising evangelicals. But when asked, missionaries would always frame their work as a response to a calling, a direct summons from God to reach out to the imprisoned. This summons was often revealed by a pastor, during prayer, in a dream or simply as a consistent feeling. This is not to say that my evangelical interlocutors were not isolated from other currents of social critique or activism. Most missionaries, regardless of where they located themselves within Brazil's racial categories, recognised and referenced anti-Black racism as a defining feature of policing and punishment in Brazil. But in our conversations, this recognition was presented as background or context, rather than an organising principle of their work. Scholars and Brazilian activists have argued that evangelical theology's combination of individualism and the universal promise of salvation provides no foundation either for a collective Black consciousness or for an anti-racist politics. 41 Others have pushed back against this argument by pointing to a growing Black Pentecostal movement that explicitly articulates Black identity through Christian theology, and vice versa. 42 In part, these debates reflect the heterogeneity of the evangelical movement as a whole. Nevertheless, this circumscribed vocabulary of race and anti-Blackness seemed to predominate in the context of my fieldwork, where missionaries almost universally framed their work as a struggle against sin and despair rather than racism.
None of this is necessarily unique to prisons. Similar discourses accompany evangelicals working in hospitals, in drug rehabilitation centres or with people experiencing homelessness. But experienced pastors and missionaries did claim that there was something particular about work in the penal system, because of the level of fortitude needed to bear the weight of suffering and to navigate the moral dilemmas faced within these institutions. Missionary work is the most visible face of this engagement. When Rio's prison administration authorises a church to provide religious assistance, the church then coordinates regular visits through the social worker stationed in a particular unit. These are limited to a maximum of six visitors. Most often, I encountered four or five women accompanied by a male pastor from their congregation. These groups act with the moral and financial backing of their church communities, and often function as mediators between incarcerated people, prison staff and outside congregations. Churches and religious NGOs also maintain various social projects that target family members of the imprisoned as well as those recently released.
What does it mean to claim that these forms of engagement constitute an 'infrastructure'? We can start with Brian Larkin's definition of infrastructures as 'matter that enable the movement of other matter. Their particular ontology lies in the fact that they are things and also the relation between things.' 43 Roads, rubbish bins and server farms have their own material forms, but they also channel or mediate systems of exchange. In many cases this matter is hidden or aspires to become so, a kind of invisible network lying underfoot that safeguards the affordances of economic and social activity. In others, infrastructure is a monumental endeavour. Latin America's grand projects of highways and dams, for example, are performances of state power, development and modernity. 44 41 See, for instance, Vilma Reis' description of evangelical churches as 'filled with Black women who cry for their children, children who fall each day from the bullets of police and from groups tolerated by police. And this is where mediatised pastors are trying to break our spine, which is called ancestality-identity-resistance.' Vilma Reis, 'Na mira do racismo institucional: Quebrando o silêncio diante da matança em Salvador', Irohin, 10: 11 (2005), pp. 10-11.
The substance and the materiality of infrastructures matter. But these physical structures are also deeply imbricated within, and penetrated by, a host of social actors. As scholars of the infrastructural turn have repeatedly asserted, it is only through this assemblage of the concrete and the social that things circulate. Beyond this general claim, there are also specific cases in which social ties act in their own right as either analogues of, or substitutes for, material channels. This observationthat there is no necessary a priori distinction between people and pipes, at least in terms of their functionallows us to reframe the productive capacity of social relationships through their ability to distribute opportunities and resources.
One notable example of this work is the concept of 'people as infrastructure', a term put forward by Abdou Maliqalim Simone that describes the constantly shifting collaborations among marginalised urban residents. 45 For Simone, these openended, often ephemeral ties nevertheless constitute an infrastructure insofar as they allow residents to consolidate livelihoods and aspirations, despite widespread institutional neglect. Julia Elyachar employs a similar approach in her analysis of Egyptian women's social networks. 46 Here, women's investments in friendships and neighbourhood acquaintances, while often dismissed by men as gossip, nevertheless constitute channels through which a variety of resources might flow. The productive capacity of women's relationships was recognised and harnessed by microfinance corporations and development programmes in order to produce a 'social infrastructure on which other projects oriented around the pursuit of profit could be constructed'. 47 In making a claim for evangelical Christianity as infrastructure, I am proposing that similar dynamics apply within Brazil's penal system. In Rio, parceria was the name given to those ties between staff and evangelicals that had been repurposed into channels. But the nature and scale of the flows bundled under that name varied considerably. Some were formalised through contracts. This was the case in the early 1990s, when the first evangelical missionaries who entered Rio's prisons negotiated agreements to fund the construction of new temples in virtually all of the state's penitentiaries. 48 More recently, the programme 'Amor que Cura' ('Love That Cures'), one of the few initiatives joining evangelicals together with Catholics, currently supplements state healthcare through a contract with administrators that allows them to cycle through Rio's prisons and offer medical attention and treatment to the incarcerated.
More often than not, however, parcerias operate as informal agreements built on the trust that churches have cultivated with administrators and wardens over many years. For example, during the period of my fieldwork in René Dotti, in one of the two men's prisons where I conducted extensive research, the photocopy machine broke down. Rather than apply for administrative funding for the replacement, the warden turned first to the missionaries working within his prisonincluding a pastor named Bruno, who later explained the details of the agreement to me. The structure of this event, in which a warden responds to a breakdown by reaching out to evangelical contacts, is almost entirely identical to Josielle's experience with the water supply. Josielle also invited me once to accompany her during a weekly visit to a juvenile detention facility 49 where, before meeting with the group of adolescents for a basketball game and a Bible lesson, she spoke with the warden to finalise the details of their plan for her church to repaint the courtyard.
I played a small role in stitching together this infrastructure, although I did not realise it at the time. In 2014, while chatting with Josielle, I mentioned in passing the Coletivo Abraçar (Embrace Collective), a network whose member organisations provided support services for people leaving prison. Josielle seemed curious, and noted down the organisation's details. The following year, when I returned to Rio and to the meetings of the Collective, I found her sitting next to Aline. The two had met here during my absence. More importantly, though, Josielle had then re-introduced Aline to a third woman, Magnólia, who Aline had first met 15 years prior. Back then, Aline was stationed as a social worker in a men's prison, while Magnólia would visit each week as a missionary. They lost contact when the former was transferred to another prison, and much had changed in the interim. Aside from her new position, Aline was now an evangelical herself, having converted from the syncretic Afro-Brazilian religion Umbanda. For her part, Magnólia had largely ceased her visits to prison and now directed a church-based NGO that offered food, training and job opportunities for both formerly incarcerated people and family members of those still imprisoned.
Over the rest of my fieldwork, I saw the encounter between the two develop into a parceria. The following year, brochures for Magnólia's NGO were taped to the walls of the Patronato and stashed in the desk drawers of technicians. By connecting with her old parceira, Aline had found a new path through which to refer clientsparticularly those looking for work, but also anyone facing food insecurity or in need of medical services. The terms of this relationship were largely like those that the Patronato cultivated with a broad variety of organisations, from secular NGOs to Catholic charities. That is, it was an oral agreement built upon personal connections and maintained through open channels of communication. But Aline preferred the churches. 'They are our first line of defence', she explained to me during one of our conversations at the end of a long working day. She had seen how government funding for civil society and the incentivisation of public−private partnerships had produced a semi-professional philanthropic class who she saw as more interested in making money than in providing any substantial help. Evangelicals, by contrast, seemed far more committed to both the social and 49 Rio's 'socio-educational' units, which imprison those under 18, are administered separately from the adult prison system, although some of these units are indistinguishable from prisons. spiritual needs that she identified in her clients. Furthermore, since 2014 many secular and Catholic groups had buckled under the weight of Brazil's economic downturn, folding or suspending most of their services. Evangelical churches were also strained by the recession, but they proved far more resilient in finding the resources to continue at least their core services. And in a moment of crisis, when the penal system itself seemed to be crumbling, people like Aline found themselves relying on these networks, and putting them to new use.

Breaking Down and Becoming Durable
It often seemed like the Patronato was falling apart. The building was not particularly old, but neither was it well-constructed, and plans for maintenance work had been put off for years. So when the roof of the men's bathroom collapsed one July afternoon, nobody seemed particularly surprised. After all, for several weeks staff had brought their own toilet paper to work because the institution could no longer stretch its budget to provide it. Months before the collapse, I spotted a leak in the water tower that stood in the middle of the staff car park, dripping steadily onto one of the cars below. A few hours later, when I passed by again as I walked to the office canteen with a correctional officer, this drip had become a steady stream. 'Operation Car Wash', the officer joked, referencing Brazil's ongoing corruption probe of the same name.
When I realised that these breakdowns were becoming more frequent, I attempted to note the changing state of the building and its facilities more systematically with each visit. My field notes from that point on are punctuated with observations about whether the office printer was out of ink again; if the air conditioning had broken; or when more blisters appeared behind the paint on the walls, the result of moisture from rainfall. This practice fostered a different kind of attention to the Patronato, one that allowed me to join in the constant talk among staff about the state of their office. As we shared our observations, these leaks, shortages and workplace hazards came to stand in for something much larger. A collapsed roof or a leaky tower were problems in themselves, but they had also become icons for the broader disintegration of the prison system, and indeed of Brazil as a whole.
The name given to this period, even as it was unfolding, was 'the crisis' (a crise). The term was a diagnosis that bundled together the downturn of Brazil's economic fortunes; a corruption probe that had snowballed into the upheaval of state and federal governments; and a rapid spike in both unemployment and recorded levels of crime. As Stuart Hall reminds us, invocations of crisis serve to unify public sentiments and anxieties in service of political campaigns to restore order. 50 In the case of Brazil, Operation Car Wash (Operação Lava Jato) served as the central justification for this campaignpoliticians wielded its narrative force to legitimise their impeachment of then-president Dilma Rousseff, while Jair Bolsonaro was fashioning himself as the icon and guarantor of a new political order. But it is important to emphasise that for the Patronato, and for Rio as a whole, crisis was also a visceral, immediate experience, one 'felt on the skin'. 52 The word gave coherence to the sudden proliferation of signs of disorder and decay that intruded upon daily life. It also furnished a narrative that could explain the sudden evaporation of a vision of national prosperity that had, until recently, underpinned individual and collective futures. When the correctional officer made a quip about Operation Car Wash, he pointed to an emerging way of reading the present, one in which a leaking water tower and a multibillion-dollar kickback scheme served as direct reflections of one another.
Of course, crisis was never the whole story. While it provided a platform for Bolsonaro's rise to power, it also offered Rio's prison administration justification for delayed salaries, withheld funds and institutional neglect. It was a ready-made explanation for why the electricity went out for almost two weeks in the training school for prison staff, or why a broken elevator forced everyone to walk up five flights of stairs to reach the administrative offices in the city centre. But the entire system had been facing similar struggles for decades. Most prisons had been built between the 1970s and early 1990s and had seen little attention since. Some of the problems were material, like flooded cells, crumbling walls and the near-constant question of the water supply. Others were legal and bureaucratic, as staff struggled to meet demands placed on them by the Law of Penal Execution. And in a system where funding priorities had always skewed heavily towards custody, treatment services continually suffered.
Crisis simultaneously turned up the volume on these problems and flattened their longer histories within a narrative of emergency. In doing so, it gave a new sense of urgency or desperation to the attempt to maintain a semblance of working order. This was particularly true of the Patronato, which was both a public-facing institution and one largely ignored by the prison administration by virtue of its non-custodial function. Not all these holes could be patched up by missionaries, volunteers or money. But as Aline and many others across the system were finding out, a surprising number could. At this point, at least, the churches were not directly repairing a collapsed roof. But they were already invested in the Patronato, an investment that seemed to spread without any clear boundaries.

The Employability Office
Between the Patronato's reception area and the men's restroom lies a short hallway, with doors on either side that lead to the various offices of custodial and treatment staff. In mid-2015 I found myself here, waiting for the occupational therapist to finish her discussion with a client so we could begin our interview. Before they were done, Aline's deputy walked past and pulled me through another door into the room designated for the small team of social workers. He wanted me to be present while he gave the good news. He had been on the phone with Pastor Lucas, one of the leaders of the Christian NGO Cidade da Graça (City of Grace). Lucas had just reached an agreement with a private bus company to hire up to 40 formerly imprisoned men as drivers in one of Rio's satellite cities. The deputy wanted the two social workers on duty to send all clients who were looking for work and who lived in that area to him directly. He would connect them with City of Grace, who would, in turn, both assist and manage their applications to the bus company.
A lot had changed by the time I returned to Rio the following year. The crisis had deepened, work was even more scarce, and Aline had been forced by the prison administration to make lay-offs. One of these was the single occupational therapist. Her office had now been transformed into the 'employability room', a space now permanently occupied and run by City of Grace. Pastor Lucas was a permanent fixture inside the office, as was another volunteer for the organisation, Raquel. While neither were technically employees, and their presence in the Patronato was entirely based on an informal arrangement with Aline, they had become integrated into the institution's rhythms, including the weekly staff meeting. Aline asked all technicians to direct those looking for work to the employability room, hoping that this redistribution of responsibilities would free up staff for other kinds of work. In turn, Pastor Lucas continued chasing down any leads for employment that he could find, and attempted to match them with the needs and skills of those who entered the office.
But not everyone was happy with this new arrangement. About a month after City of Grace arrived, I found myself in the social workers' office again, this time with Katia, while she took a short break from that day's stream of orientations. While we sat together, she began discussing her growing misgivings with the Patronato as a whole. It was a little surprising to me, since we had previously discussed how much she enjoyed working there. This day her attitude seemed completely different. She was frustrated with her workplace, and more specifically with the two City of Grace volunteers who sat in the office across the hall.
First, she was concerned that even after a few months, 'I'm still not really sure what they do here'. This statement carried an undertone of suspicion regarding the motives of the volunteers, one that I will return to below. But it was also fairly clear at that point that employment was not the only, or even the main, concern of the employability room. For a while now, both volunteers had moved beyond their original remit. Raquel, for instance, was a lawyer. Soon after she arrived, she began offering counsel to any clients with outstanding issues related to their sentences. In addition, she spent a considerable amount of her time both inside and out of the office searching for financing opportunities. As she explained to me during an interview: Raquel: [We can get] these funds through the NGO, there are some special groups actually within the Justice Tribunal in relation to restrictive penalities, and these funds can be made available by the Tribunal to the private sector. And nobody's using it. Because the most important thing is for us to find a funding agency to reform the Patronato, reform the Patronato so that it can actually have an impact. David: What do you mean by reforms?
Raquel: The Patronato just needs materials, because the apenados ['convicted'] can provide the labour power, so what we need to begin is construction material, to make a second story for the building, to modernise the offices, the archives, we need tables. So anyway, the first step is for construction materials and office supplies, basically, everything that we've had difficulty with. Because here in Brazil the vision businesspeople have of the egresso [the recently released], the apenado, it's hard, but we'll try, right?
In this short exchange, Raquel identified two distinct strategies that City of Grace was pursuing to help fund the institution. The first took advantage of its status as an NGO to receive federal funding, which could then indirectly be diverted back to the Patronato. The second was to find private money for the Patronato through business or church donations for basic maintenance and repair work. In both cases, City of Grace would become a node to funnel private funds into a state organisation and to mediate the relationship between the two. But Katia's concerns went beyond this question of mission drift. More worrying to the social worker was that volunteers had been given access to the archives, a small room containing the case files of each client. This access was related to yet another new branch of City of Grace's work: calling those who failed to attend their appointments to encourage them to return, and reminding them that they could face rearrest if they failed to show up. For Katia, this was an affront to the professional confidentiality (sigilo profissional) required of her occupation. She began to self-censor her reports, out of concern that the volunteers could leak sensitive data. This would limit the information that could be used to 'accompany' the client, but it would also protect them from harm.
Katia resigned from the Patronato two months after this interview. I cannot say to what extent her misgivings about City of Grace influenced her decision. She, like everyone else in the office, was frustrated with her growing workload, an unresponsive prison administration and the two months' salary she was owed at that point. But since there was no money coming from the administration, nobody would be hired as her replacement. At the same time, Raquel was growing more comfortable in the office: In the beginning we were a little off-putting, because the functionaries, they thought that we were a threat to their jobs. I don't really get it, but there was a kind of restriction, there was no parceria. But now that we're taken seriously, now that we're established, they're directing egressos to us for work, social workers are directing them to us, the psychologists too, so it's starting to happen. They're starting to appreciate our work, to understand the value of our work.

Push and Pull
In June 2015, Rio's Núcleo de Defesa dos Direitos Humanos (Human-Rights Defence Nucleus, NDDH) released a report detailing the results of two official visits to Evaristo de Moraes, a maximum-security or 'closed' men's prison. 53 In keeping with NDDH policy, the warden was not advised beforehand of either visit. The report describes an isolation cell 'flooded by a blocked "toilet", with water rising up to 10 cm from the floor'. 54 After the visitors brought the issue to the attention of the warden, the latter transferred those currently held in this cell to other areas of the prison. But on their return, they found that it was once again occupied and in the same condition as before. NDDH members also noted that there appeared to be no light coming from the two cells neighbouring this one and asked the warden for an explanation. Initially, he said that those held in the cell had unscrewed and sold the bulbs; but when the doors were opened, it was clear that they had simply burned out. Switching tactics, the warden then argued that the prison did not have the resources to replace themuntil one of the visitors reminded him that federal legislation prohibited unlit cells, after which two replacement bulbs quickly appeared.
Not all cells were in the same state. The NDDH report also noted in bold that 'some cells (especially those destined for incarcerated evangelicals) are in much better condition'. 55 These are better known as the 'evangelical cells' (celas evangélicas) and are ubiquitous across all of Rio's prisons, as well as much of Brazil and Latin America. They began to appear in Rio in the mid-1990s, after incarcerated people and missionaries campaigned for separate spaces where evangelicals could live in communion with one another, in accordance with their faith, and with a selfimposed code of conduct. 56 Over time, the broader imprisoned population came to name the cells not designated as such the cells of the 'ímpios' ('faithless' or 'impious'). 57 The contrast between the two is often stark. During my fieldwork in René Dotti, I regularly passed by the cells of the impious, followed by those of the evangelicals, on my way to the temple at the far end of the prison. While I never entered any cells during my research, the difference was visible through the bars: evangelical cells were cleaner and better maintained. They often seemed literally brighter.
In part, this disparity stemmed from the social order maintained in evangelical cells, an order that included a regimented daily cleaning schedule. But it was open to other interpretations. Once, as I walked through the wings of René Dotti with a missionary during her first visit to a prison, she asked the rhetorical questionwhy would anyone choose to stay in the cells of the impious? For her, our journey represented a clear before-and-after narrative through which we had witnessed the saving grace of faith. My reaction was more cynical. I understood the disparity as the result of a steady stream of donations of food, clothing, bedding and hygiene/ 53 Defensoria Pública do Estado do Rio de Janeiro, Núcleo de Defesa dos Direitos Humanos, 'Relatório de visita à unidade prisional ', doc. EE-20/001/1684/15, 9 June 2015, avaliable at www.defensoria.rj.def.br/ uploads/arquivos/4489b8a1ce11422d9e26ab69ec85a514.pdf, last access 5 April 2022 Ibid., p. 10. Here the quotation marks cast doubt that a hole in the ground could be properly called a toilet. 55 Ibid., p. 36. 56 For accounts of these regimes of segregation, see Dias, A igreja como refúgio; Johnson, If I Give My Soul.

57
Despite the name, many of those imprisoned in these cells are religious; this includes devout evangelicals who simply do not wish to be placed in the evangelical cells.
cleaning materials that missionaries funnelled into evangelical cells. The NDDH report seems to insinuate this as well, although it also carried a much broader accusation, if only implied: that this was a clear case of favouritism not just on the part of missionaries but also prison staff, including the warden.
While the report directly addresses the conditions of a single prison, it speaks to more general forms of inequality that have proliferated as evangelical missionaries, and the resources they bring, become integrated within the processes of imprisonment. Evangelical Christianity as infrastructure shapes which cells are left flooded and which are maintained; where the light shines and where it is dark. More broadly, evangelicals take part in a kind of triage that determines who within prisons is given access to goods and services, and through what routes. To borrow from the language of urban infrastructures, they have a 'splintering' effect, fragmenting the experiences of confinement that are already shaped by various other faultlines. 58 As mentioned above, this effect emerges from the artificial scarcity imposed by the prison administration far more than any deliberate plan on the part of missionaries or others. But even when donations or social programmes do not specifically target evangelical cells over others, the former are almost inevitably prioritised.
At the same time, evangelical Christianity has brought a series of displacements to the penal system, and has contributed to a broader hostility towards other religious agents. Notably, this includes discrimination against Afro-Brazilian religious traditions and practitioners. During an interview, the mãe-de-santo 59 of an Umbanda congregation explained that she was forced to hold her services in the prison cafeteria rather than the unit's temple: 'If I could, I'd use the same temple that the law promises me. It would be easier. But it was made clear to me that it wouldn't be wise to even ask because I'd end up fighting with the pastors.' Her experience is not isolated; it forms part of a broader landscape of religious persecution in Rio and Brazil, where the hostility of many evangelicals towards Umbanda and Candomblé often erupts in violent attacks against their adherents, icons and sites. 60 Tensions also emerge within the evangelical tradition itself. During my fieldwork, an openly gay pastor belonging to an 'inclusive' church began weekly visits. 61 But each week he faced barriers to entry, including guards who would deliberately leave him waiting outside the prison; missionaries who would harass him as he passed on his way in and out; and imprisoned evangelicals who would attempt to bar him and any 'homossexuais' from entering the temple. His visits stopped after six months.
Like The term refers to a religious leader within Afro-Brazilian religions but literally translates as 'mother-of-the-saint'. frustration, it was more often expressed indirectly, through awkward pauses and sentences that seemed to trail off, leaving me to complete the accusation. It was also often mixed with respect, since most recognised that evangelicals were among the only groups actively interested in supporting the well-being of those caught in the gears of criminal justice. Nevertheless, the concern that evangelical missionaries and staff were taking part in a concerted 'push' to exert control was never far from the surface.
This accusation is not new, nor is it only voiced by prison staff. In fact, it finds echoes in the work of anthropologist Rita Segato, who wrote on her experiences building an education programme within a Brasília prison in the 1990s. 62 Reflecting on the ongoing tensions between her work and that of missionaries, Segato argues that evangelicals (and, to some extent, charismatic Catholics) claimed for themselves a 'monopoly' on redemption and positioned themselves as its gatekeepers. But she also signalled that these groups would 'take part in the administration of the carceral communitythat is, more than simply developing religious work, they take charge of the organisation of all daily activities, administrating the provision of services for both those who are converted and those who are not'. 63 Missionaries were aware of these critiques. In fact, it was Magnólia who introduced me to Segato's text, to gauge my reaction and to explain the potential harm that an anthropologist might cause. But these arguments assume that evangelicals are willing participants who smuggle in particular spiritual or political projects. As such, they gloss over or entirely ignore how missionaries, like many others, are pulled into the various roles that they play. That is, while evangelicals might repurpose the institutions of punishment, the inverse is also true. Most of the pastors, missionaries and volunteers with whom I worked had a highly developed sense of self-critique, and voiced their discomfort with the ever-expanding scope of issues they were expected to solve. While some, like Raquel, took pride in their work, others, like Josielle, would not forget that they had been coerced into taking on the responsibility of the state. For the latter group, this work was a distraction. In another interview, a pastor explained her frustrations with organising donations of clothing for a state women's prison, claiming that 'none of this is our responsibility, this isn't what we are here to do. What we should be focussing on is-' and finishing her thought by pointing towards the sky.
When Josielle spoke of her conflicting feelings with me, she turned to the Bible passage that detailed the miracle of Jesus feeding the multitude with loaves and fish. What she took from the event was that the work of evangelising was always 'integral', that material support was always woven into spiritual care. This did not blunt her criticism of the prison administration, but, for her, it offered a sense of purpose within a position in which she had somewhat unwittingly found herself. She, like many other volunteers, sensed that she was bound by her parcerias, and by the necessity of maintaining good working relationships with the prison system. But she also keenly felt a duty to preserve the dignity and survival of those in prison as part of her own calling. There was often a deep sense of fulfilment that came 62 Segato, 'Religião, vida carcerária e direitos humanos ', p. 140. 63 Ibid., p. 142. with this work. When I spoke with Raquel one afternoon at the Patronato, she explained that although she earned no money from her time in the institution, she did it 'for love'. Aline echoed her thoughts a few weeks later when she explained that 'Pastor Lucas gets a minimum wage [from the NGO] and Raquel gets nothing. They do it for love.' At this point Aline hesitated, thinking through her words carefully before she continued. 'It's like a division of responsibility, which is hard for me to share with you because it should be my responsibility, but in reality it just isn't, it's ours.' If it was love that brought Josielle and Raquel into the penal system, this love had been refashioned as a pump, one that could draw water, money, construction materials and job opportunities. And it was becoming increasingly difficult to conceive of governing these institutions without it.

Conclusions
Aline once spoke to me of her plans to build a prayer tent in the Patronato's car park, where volunteers from any church could offer guidance to clients. The project came from her recognition that technical staff were simply not able to offer clients what they most desperately needed: forgiveness. In the end, this tent was never built. But the idea was only imaginable at the point where an institutional crisis, a set of established parcerias and a director's religious and professional convictions converged.
Others had their own projects. Pastor Bruno, who worked in René Dotti, had made plans with the warden to transform it into Brazil's first evangelical prison. 64 Their vision was of a penitentiary that would exclusively imprison members of the faithful and, in doing so, provide them with sanctuary. For her part, Josielle had begun to search for new platforms to advocate for less church involvement and greater state responsibility in the provision of essential services. But all this work was interrupted, and the terms of engagement between churches and the penal system scrambled, by the arrival of the Covid-19 pandemic. In March 2020, all missionaries were banned from entering prison. Many shifted their efforts to supporting prevention and treatment efforts, especially through the supply of basic health and hygiene equipment. Amor que Cura also financed and managed the installation of a new tele-health system in various prisons across the state. Meanwhile, the Patronato's offices were largely emptied out as those on parole and house arrest were given temporary reprieve from the obligation to attend.
As these examples show, evangelical Christianity as infrastructure does not refer to a stable or unified set of relationships. It is the product of ongoing and fragmented transformations that are collectively, in fits and starts, integrating Christian networks into penal governance. Evangelical Christianity and mass incarceration have emerged together in the last 30 years, shaping one another in the process. Faith and necessity bind staff to church groups as they work to resolve, or at least mitigate, the strains of an institution that always seems on the verge of falling apart even as its expansion continues unabated. But this repair work also represents a critical shift: where evangelical networks have proven their flexibility and strength in reaching those outside 64 They had initially thought it would be the first in Latin America, until Pastor Bruno learned that one was already operating in Chile. the state's grasp, a series of infrastructural 'crises' now re-route this carrying capacity towards propping up a state institution. The growing precarity of this carceral project tethers two central concerns of missionariesthe survival and the salvation of the imprisonedto 'saving' the prison system itself. A discourse of parcerias might smooth over the tensions, but this is also a hostage situation.
These phenomena are not confined to the boundaries of Rio de Janeiro or Brazil. The work of evangelical missionaries, staff and NGOs has reshaped the landscape of punishment across Latin America, and will continue to do so. And while much of the scholarship on the region's prisons is framed in contradistinction to the North Atlantic, there are also clear parallels with the United States, where evangelical involvement arguably stretches back much further. We might compare, for instance, the case of Rio de Janeiro to that of Florida, and specifically to the Salvation Army's history of involvement with punishment to the point where, in the 1970s, it assumed responsibility for the entire state's probation services. 65 As Cyrus O'Brien notes, this partnership, itself the first modern privatisation of probation, was: a marriage of convenience. And the financial incentives involved in monitoring tens of thousands of Americans under court supervision made divorce difficult or impossible. The Salvation Army's mission 'to preach the gospel of Jesus Christ and to meet human needs in His name without discrimination' carried the organization into urban areas. The financial and coercive capacities of the state pulled the Salvation Army into profitable systems of punishment. 66 The economics of these cases differ, but they converge in the dynamics of churches and systems of punishment repurposing one another. In drawing such cases together, we might better understand the forces that both push and pull evangelical groups into the service of punishment across the Americas.
Here I have largely approached incarceration from the perspective of prison staff, missionaries and volunteers. Such an approach leaves unaddressed the implications of evangelical Christianity as infrastructure for the imprisoned, those on parole, their families and their communities, a task of critical importance for future research. For now, we might speculate that for these groups, some shifts are barely felt, like when church groups fund a new copy machine. Taken collectively, these processes still do not seem to disturb the consolidation of Brazil's penal system, nor the legitimacy of this carceral projectat least, not yet. But when the taps run dry, the lights go out or an infection spreads, evangelical Christianity as infrastructure becomes a question of life and death. And this is true not (or not only) because evangelicals take part in regimes of shared governance, but rather because they have been recruited into maintaining or replacing systems that make both governing prisons and surviving them possible. This is what thinking with infrastructure reveals: that the parcerias binding missionaries, churches and NGOs to the prison system are also channels that underwrite Brazil's carceral project. The long-term implications of these partnerships are unclear; their limits continue to be tested by prison administrators and state governments, who search for new pathways to redistribute responsibilities and to source labour or materials. Aline, Josielle, Katia and the NDDH have their own misgivings about such shifts, even as they participate in them. But these experiments in infrastructure are, in themselves, a reworking of punishment in Rio de Janeiro and Brazil, leveraging faith and solidarity to draw new investments into the project of incarceration.