Dealing with violence: Varied reactions from frontline workers acting in highly vulnerable territories

Introduction

This paper analyzes how frontline workers, as professionals responsible for implementing governmental policies, deal with daily violence in highly vulnerable territories. Here, we take into account that the State relates in multiple ways with social violence, understood as an interpersonal phenomenon (Blume, 1995; Reiss and Roth, 1993) with specific spatial dynamics (Wacquant et al., 2014).

The relationship between violence and the State is a traditional issue in the social sciences' classic literature (Bourdieu, 1980, 1988; Elias, 1939/2000; North, 2009; Weber, 1919). In dialogue with Weber (1919), almost all these authors argued that the modern State could never exercise a monopoly of violence. While the State is the prime holder of legitimate violence, some illegitimate violence always remains dispersed in society and would be used by social agents in different ways, and ascribed different meanings (Bourdieu, 1980; Elias, 1939/2000; North, 2009; Sundström, 2016).

Contemporary studies, inheriting the problem from classic sources, have been trying to understand how the State deals with the monitoring of its monopoly on legitimate violence (Johansen, 2019; Kyed, 2019; Morin, 2013; Pauschinger, 2019; Reid-Henry, 2015; Rivera and Ward, 2017) and with challenges in controlling social violence. This includes research on borders and their intersection with violence and conflict (Brambilla and Jones, 2019; Jones, 2015; Schindel, 2019; Weber, 2012), the relationship between institutions, violence, and crime (Barnes, 2017; Vigneswaran, 2014), narco-traffic (Ballvé, 2012; Warburg and Jensen, 2018), and terrorism (Ezcurra, 2017). Such literature usually analyzes violence as a general dynamic that affects the nation-state or society as a whole.

However, we also know that violence—and ways to deal with it—manifest differently across places and among social groups (Duru, 2019; Keenan, 2019; Kyed, 2019; Loftus, 2015; Mawani, 2012; Murphy et al., 2018; Pilo, 2019; Shaw, 2019; Vigneswaran, 2014). Violence, after all, is a social and interpersonal phenomenon (Blume, 1995; Reiss and Roth, 1993), which materializes in daily power relationships among actors operating within specific territorial dynamics, struggling over questions of resources distribution, conflicts of interest, and freedom of action (North, 2009). When manifestations of violence are only observed as one broad process, we are unable to understand why and how violence manifests in various ways among different people and territories. To advance the understanding of such situations, we should be able to observe the micro-dynamics of the relationship between State and violence that happen in specific local contexts.

This article aims to contribute to the literature by observing these micro-dynamics in one type of territory marked simultaneously by inequality, high levels of violence and exclusion. Previous studies have already highlighted the relationship between violence and excluded social groups (Reiss and Roth, 1993). As they act at the State’s borders and are the face of the State for citizens, frontline workers are daily exposed to multiple situations of violence in the places where they work (Sundström, 2016).

We consider that (1) the State is a set of heterogeneous actors and institutions, and not a monolithic power (Sharma and Gupta, 2009); (2) that it has to dispute its monopoly of violence with other actors (Elias, 1939/2000; North, 2009); and (3) that it does so within specific territories in which conflicts emerge (Pinto and Do Carmo, 2016; Wacquant et al., 2014; Zittoun, 2014). Applying these definitions to the frontline workers, while some of them are expected to deal directly with situations of violence, such as police officers (Epp et al., 2017; Nicholson-Croty et al., 2017) and those tackling issues such as violence against children or women (Frattaroli and Teret, 2006; Hong, 2016; Levenson and D’Amora, 2007; Morin, 2013; Pauschinger, 2019; Riccucci et al., 2014; Rivera and Ward, 2017). Other workers who implement policies not directly related to violence can be affected by it, having to deal with situations of criminality or violence.

Only a few studies have contributed to mapping these processes of everyday violence against frontline workers. When they do it, frontline workers are seen as victims or passive agents in the face of violence, whose usual strategy is to escape or avoid such situations. In this paper, we argue that the range of reactions observed among workers exposed to situations of violence is very complex. We demonstrate that their repertory cannot be reduced to passive behaviors: frontline workers can also use their agency to proactively deal with violence. Some of them do accept violence as part of their working context, maintaining silence about it or giving up on implementing policies, as shown in the existing literature; others try to adapt policies and negotiate with the dynamics of violence; and yet others do not take violence for granted and transform their policies into strategies against violence.

Thus, in this paper, we consider that a complete explanation of the relationship between violence and frontline workers should involve three aspects: an analysis of the various ways in which violence manifests itself; a typology of the professionals’ reactions in the face of violence; and an understanding of the reasons why agents subjected to these situations of violence adopt different positions and strategies.

Here, besides analyzing the diverse manifestations of social violence and agents’ reactions to it, we analyze how the territory plays a role in explaining different reactions. We demonstrate that workers’ ability to develop different reactions relies on the connections they and the policies have with places. When policies and agents are territorially embedded, they are able to react to violence both adapting to it or facing it. And when policies and agents are territorially blind, there are no institutional devices (under the policy’s angle) or social ties (under the agent's angle) that can be mobilized by them. Consequently, their reactions are about silence to violence. Territorial embeddedness means policies and actors are connected to territorial dynamics of power and mutual recognition, being able to dialogue with local needs; while territorial blindness refers to policies and actors disconnected to the local dynamics.

The empirical evidence is composed of cases in which frontline workers who do not work directly with criminality issues are regularly exposed to violence in their work. We gathered data from interviews of 140 frontline workers acting in different policy areas (community health workers, nurses, teachers, and social workers), who operate in very vulnerable and violent neighborhoods of Sao Paulo. The conclusions obtained are relevant to a variety of contexts because the cases analyzed reproduce conditions very similar to those found in several other developing countries and vulnerable territories in developed countries, also marked by high inequality, vulnerability, violence, and significant crime rates.

The paper is organized as follows. In the first section we present the literature about violence and the State, and the consequences for frontline workers. In the second section we comment on the context, data, and methods. In the third section we present the analysis of the multiple reactions against violence developed by frontline workers and a typology of strategies available to deal with violence. In the fourth section, we discuss the findings and propose a patterns of reaction taxonomy and their relationship to territories. The conclusion points out the main consequences of the research findings and future research agenda.

Violence and the state: Issues for frontline workers

Violence is a worldwide and multifaceted phenomenon. The fight against violence has always been one of the most important governments' agendas since the constitution of the first modern States (based on the idea of concentrating military power) until the present day. The specialized literature on violence has focused on studying this broad phenomenon of how States protect their territories and citizens, and fight crime (Brambilla and Jones, 2019; Bond and Gebo, 2014; Ezcurra, 2017; Marenin, 2014; Pinto and Do Carmo, 2016; Weber, 2012). Some studies have also analyzed specific policies against everyday violence, such as policing practices and social work initiatives to protect women, children, and vulnerable groups (Epp et al., 2017; Frattaroli, 2006; Johansen, 2019; Kyed, 2019; Levenson and D’Amora, 2007; Meier and Nicholson-Crotty, 2006; Nicholson-Croty et al., 2017; van Steden et al., 2015). Other researchers have studied violence as a broad social and political phenomenon (Cover, 1986; Fanon, 1963; Foucault, 1977), in which power takes on different meanings: some authors have introduced a distinction between power and violence (Arendt, 1970), while others have described an intrinsic relationship between these concepts, with violence assuming a role in the production and maintenance of power relations and oppression (Foucault, 1977) and even a cathartic or revolutionary character (Fanon, 1963).

In this article, inspired by WHO (2002) and Reiss and Roth (1993), we consider violence as the intentional use or threat of physical force or power by persons against persons that may be expressed in different ways, as injury, death, physical or psychological harm, maldevelopment, or deprivation. Although violence is an indispensable constitutive phenomenon of human nature and, therefore, of any social relationship, the uses of violence by the exercise of power can occur in different ways (Bourdieu, 2012; Elias, 1939/2000; Žižek, 2009). What determines these different ways of exercising violence can be aspects associated with the specificities of how the civilizing process takes place in each social configuration (Elias, 1939/2000), with the characteristics of the field or domain of reality that is taken into account (Bourdieu, 1980), among other factors.

Therefore, even if violence manifests itself as a widespread phenomenon, it occurs in very different ways in different places and is directed against different social groups (Keenan, 2019; Kyed, 2019; Murphy et al., 2018; Pilo, 2019; Pinto and Do Carmo, 2016; Vigneswaran, 2014). Studies have shown how some groups are more subject to situations of violence, mainly based on their race, gender, ethnicity, religion, or neighborhood (Epp et al., 2017; Mawani, 2012; Nicholson-Croty et al., 2017; Reiss and Roth, 1993; Wade-Olson, 2019). Research has also shown that cases of violence are more concentrated in some places than in others (Duru, 2019; Loftus, 2015; Shaw, 2019) and are more heavily clustered in extreme contexts, such as in poor or very unequal countries (OECD, 2019). This means that violence can be considered a social process caused by conflicts between actors operating within specific territorial dynamics.

To understand such dynamics, it is essential to recapture that territory is a polysemic concept, including various definitions in different theories and disciplinary traditions (Hasbaert, 2004). A territory can be defined by administrative or legal-political boundaries (Ratzel, 1990); cultural and identity aspects (García, 1976; Hall, 1986); or economic basis (Godelier, 1984). All definitions share a common substratum: the exercise of power in the dispute for control of material or symbolic resources between members of a territory, or between them and those outside (Raffestin, 1980). Although the key variable about territory is power, it can be operated from a structuralist perspective, according to which it is only important to recognize how power reproduces itself in this spatial unit. Or it can allow exploring the interdependencies between structure and agency, involving the structure and change of the local social orders; that is, recognizing the constraints that limit the freedom of agents and how they use this small margin of action that remains open to act on these structures. ¹

North et al. (2009, 2012) establish a clear distinction between social orders in which the monopolistic concentration of legitimate violence in the hands of the State prevails and those in which the incomplete centralization of the use of force generates a dynamic marked by high levels of economic inequality, asymmetric power, and constant threat to the physical and social integrity of people and social groups. Although those authors associate social orders at the scale of nation-states, other authors apply similar explanatory logic to smaller-scale territories, such as rural places in Latin America (Berdegué et al., 2015; Favareto et al., 2015), neighborhoods on the outskirts of large cities and metropolises in the USA (Wacquant, 2007) or peripheral regions in Italy (Felice, 2013).

Other studies on vulnerable territories in cities in Latin America, Asia, and Africa offer similar findings, suggesting that, in these places, the frontiers between the world of the crime and the world of the State are quite blurred (Barnes, 2017; Feltran, 2008; Fontes, 2019; Sanjurjo and Feltran, 2015). In these territories, crime and State compose legitimate orders producing a specific device of urban organization due to their tensions and accommodations – a legitimation quite different from that suggested by the Weberian conception, but of concrete effectiveness for people who live in places where the State is absent or has a fragile and ambiguous presence. In this way, criminal organizations may develop collaborative and competitive arrangements with the State to determine acceptable levels of violence and stability (Barnes, 2017). These studies suggest that the micro-dynamics of the relationship between State and violence may create many types of reactions among frontline workers.

Considering that policies are influenced by the spaces in which they are materialized (Pollitt, 2013), we can presume that, in vulnerable territories with high levels of violence, the State and its agents are exposed to situations of violence even when implementing policies not designed to fight against violence. What remains unclear, then, is (1) how violence affects the State and its agents even concerning policies that are not supposed to face violence directly, and (2) how the State and its agents deal with these situations of violence in the territories in which they manifest.

The answers to these questions presuppose looking at the micro-dynamics of policy implementation in vulnerable territories. They also presuppose observing how workers at the frontline deal with such situations of violence and make decisions. The literature has already analyzed some of the effects of violent situations on frontline workers’ behavior. These studies show that some workers in certain professions are exposed to many different types of violence, such as verbal aggression, threats, and physical assault (Koritsas et al., 2010; McMahon et al., 2014; Robson et al., 2014; Sundström, 2016). Some studies term such instances as workplace violence (Koritsas et al., 2010), when the place in which workers implement policies is strongly associated with situations of violence from clients.

Most research, however, focuses on the violence provoked by clients and not on the territories in which violence is a part of the context. Such research usually concludes that the effects of violence in policy implementation involve frontline workers developing coping strategies, becoming subject to mental illness, or giving up service (Diamond, 1997; Koritsas et al., 2010; McMahon et al., 2014; Robson et al., 2014; Tummers et al., 2015; Zang, 2017). Some workers react passively, trying to avoid the violent situations, while others develop certain types of learning for dealing with violent situations (Robson et al., 2014). Overall, these studies usually view violence solely as an obstacle to policy implementation. This is because they usually consider that frontline workers are state agents trying to implement and legitimize the State’s interests. In this paper, however, we do not consider that frontline workers act only as implementing agents.

Frontline workers have agency, which is the ability to deal routinely with the real situations they meet (Maynard-Moody and Musheno, 2003), forming judgments and taking actions (Maynard-Moody and Musheno, 2012), and doing what they can do (Brodkin, 1997). Agency is shaped by their position and status in society and in the territory and by their authority (Giddens, 1979; Sewell, 1992). At the same time, the agency is expressed in interactions with and against citizens (Maynard-Moody and Musheno, 2012), which are always embedded in specific territories and social dynamics. Comprehending the connection between agents and the territories where they work is an important element to understand how they use their agency to interact with citizens (Siblot, 2006). Therefore, to understand how frontline workers deal with violence, we have to observe the multiple forms of reactions based on how they use their agency as actors with social positions in specific territories. In the subsequent pages, we contribute to the literature by proposing a taxonomy of possible frontline workers' reactions when dealing with multiple situations of violence during policy implementation, going beyond the behaviors usually identified.

Context, data and methods

We intend here to analyze the daily relationship between the State and violence, trying to understand how street-level workers deal with various situations of violence within vulnerable territories. To do so, we focus on frontline workers that implement social policies in vulnerable neighborhoods in the city of Sao Paulo in Brazil, one of the most unequal and violent countries in the world (Lima et al., 2016; UNDP, 2019). Only in 2018, Brazil had 57,341 homicides and 73,000 reported cases of violence against women (IPEA, 2019), just to cite some numbers.

Sao Paulo, the largest city in the country, has many vulnerable neighborhoods that witnessed a considerable increase in urban crime in the 1990s, reaching a peak of homicides at the end of the 1990s (Fontes, 2019; Manso, 2012). While homicide levels have fallen more recently, shoplifting and robbery rates have been increasing constantly (Fontes, 2019). The city has also seen high levels of violence against women and drug use disorder. Many of its vulnerable territories are dominated by paramilitary groups and drug dealers (Dias and Manso, 2019; Feltran, 2019). Thus, the case of frontline workers in Sao Paulo can be seen as an extreme case, which is a way to reach a “hard-to-get-at organizational phenomena” (Hällgren et al., 2017: 112).

The data collected for this paper are part of a broader study that aims to analyze the practices of frontline workers implementing universal policies in areas with high levels of violence, vulnerabilities, and inequalities. We selected the four neighborhoods with higher homicide rates, higher Gini indexes, and the lowest human development numbers in the city. For the larger project, we interviewed 173 frontline workers working in different policy areas in these neighborhoods: 93 community health workers, 20 nurses, 40 teachers, and 20 social workers. In this paper, we analyze the interviews of 140 respondents who voluntarily reported situations of violence when responding to certain questions in the interviews. The interviews consisted of three parts: 1) profile and trajectory, 2) perceptions about clients and policy implementation, and 3) stories of success and failure. The interviews were conducted between 2017 and 2019.

These workers implement different policies: nurses and community health workers implement the family health strategy, a primary health policy aimed at improving health conditions based on prevention practices. This policy is implemented by teams based in health clinics, which are responsible for one specific geographical area. The teams must follow up on health conditions through both house visits and services offered in the clinics. Teachers work in public schools at both the primary and high-school levels. In Brazil, children from lower-income families usually attend public schools, while middle- and upper-class families often study in private schools. Public schools in Brazil face many problems, including a lack of resources, low-paid and less educated teachers, and high levels of violence against teachers inside schools. Not surprisingly, then, Brazil still has one of the highest levels of inequality in education (PISA, 2019). The social workers interviewed work in centers that offer supplementary services for children and teenagers living in vulnerability. These centers conduct several activities during non-school hours. They help keep children off the street, take care of them while their parents are away at work, and help them strengthen their social and emotional skills.

All interviews were transcribed verbatim and analyzed using the N-Vivo qualitative data analysis software. Initially, each interview was coded separately (Nag and Gioia, 2012) two times by different authors. After that, we compared the coding, and when we found discrepancies, we reconciled them in the coding book. During this process, one of the codes that appeared recurrently was “violence”, which was coded in 140 of the interviews. Considering the centrality of this dimension, we decided to code the material again, only looking for descriptions of violence. In the next step of coding, we tried to find different types of reactions that appeared in the material. We did this procedure until attaining theoretical saturation of the material with first-order codes (Glaser and Strauss, 2017). We found eight first-order codes in this phase. After that, we grouped the first-order codes into second-order codes, which gave us three types of patterns for dealing with violence in policy implementation. At last, we observed how these patterns relate to territories.

The multiple reactions to violence

When implementing policies in highly vulnerable neighborhoods, frontline workers face a variety of everyday manifestations of violence. Issues such as drug use disorder, violence against women and children, drug trafficking, assault, and other types of violence present themselves to workers as conditions of the territory in which they work. As some of the interviewees reported, violence is a part of the context, and they are obligated to deal with it, even if the policy they are implementing is not designed to do so. When analyzing how frontline workers deal with these multiple manifestations of violence, we found eight basic types of reactions that represent the ways they use their agency, which are described below. Far beyond taking a passive posture towards violence or considering it a limitation of their work, we found that frontline workers develop various strategies and reactions when they face violence.

Discussion

The eight reactions described in the previous session show how frontline workers adopt multiple attitudes when dealing with violence. Far from being state agents that only implement policies, data show that these frontline workers develop their own reactions to deal with the situations they find in the specific territories where they work. In these territories, their responses to violence reflect their agency in dealing with these situations and making policy every day from the bottom up (Maynard-Moody and Musheno, 2003).

From the data collected, we were not able to associate individual reactions with particular professions or identify individual workers with one specific strategy. Notwithstanding, it is possible to observe that the reactions are associated with how the policies and the frontline workers are connected to the territory, understood not only as a given spatially determined area, but mainly as the place where power is disputed (Raffestin, 1980) in its material and symbolic dimensions. To observe this relationship, we aggregated the types of reactions to see the mechanisms frontline workers develop when dealing with situations of violence. We found three patterns.

The first involves accepting the situation and surrendering to it. This occurs when workers maintain silence about situations of violence, deny the effectiveness of a policy, or give up working on it. In our case, violence may indeed become an obstacle to policy and frontline workers. These are the kinds of situations that have been more extensively studied in the existing literature, showing how difficult situations make workers exit or neglect their role. They also reflect cases in which frontline workers feel powerless, resourceless or unable to deal with situations and, as a consequence, may cause the State to lose its power to implement policy in that area.

The second pattern involves accepting the violence but adapting policies and actions to get things done. This happens when frontline workers exclude specific cases, when they adapt their actions, or when they negotiate with criminal elements. In all these cases, workers are not fighting against violence; they accept it as part of the context. At the same time, they do not surrender, they try to enable policies in whatever ways they can. These are situations in which the violent conditions of a territory alter policy. Frontline workers adapt the original plans to real conditions, even when that means negotiating with illegality. This mechanism shows the blurred frontier between the State and crime already demonstrated in some studies that analyze violent contexts (Feltran, 2008).

The third pattern involves non-acceptance and confrontation of violence. This happens when frontline workers do not take the violence for granted and fight against the violence and its failures. They decide that the situation is not acceptable and try to intervene in it. In doing so, they change the policy's content and meaning, transforming it into a confrontation of violence. Sometimes they are able to do it; sometimes, they feel frustrated about it. However, in both situations, they assume that their main purpose is to face a violent situation and solve it. This mechanism makes frontline workers use their agency to reinforce the State's role in dealing with violence, even in situations not designed to do so. In this way, it reinforces their proactive role and, at the same time, reinforces the frontiers between the world of crime and the world of the State through a refusal to accept illegality and violence.

These patterns and the related reactions are described in Figure 1.

OPEN IN VIEWER

Figure 1.

Reactions to and mechanisms against violence.

The question that remains is: how are these reactions and mechanisms connected to the territories? When policies and agents are not embedded in the territory, including their power relationships and disputes, their reaction is to accept and surrender. This is because they do not understand the local dynamics, they do not have the needed assets nor the embedded knowledge, resources or social network with local stakeholders to include, in their strategies, ways to face the violence. This is a phenomenon in which policies or agents are territorially blinded. The first case is when the policy is designed not to have a direct connection to the territory. This means the services are delivered there as they could be delivered anywhere else. It is the extreme case of schools when they do not have practices involving local communities or territorial adaptation of institutional components of pedagogical practices. The second case is when frontline workers are also disconnected from the territories. This may happen when the workers are new in the services and do not know the territory and the persons; or when they work in different places and do not connect themselves to them, as they stay only for short periods in each place. Consequently, they do not know local repertoires and disputes, the local language, or connections that could help them in their work.

On the other side, when policies and agents are embedded in the territories, their reactions may be adapting or confronting violence. In both cases, the ability to adapt or to confront depends on the connections between policies, agents, and territories. This is because these mechanisms rely on the capacity to activate local knowledge, resources and assets only available in situations of policy and agents’ embeddedness. We saw these mechanisms when policies are implemented incorporating local elements, as citizens’ social networks and dynamics. Or when frontline workers have a strong tie to the territory – they live there, they have local ties, and so on., which means that they can use their social connections into the policy, as previously shown in the case of health policies (Lotta and Marques, 2020). The different reactions to violence depend, after all, on the degree of policies and agents’ territorial blindness or embeddedness.

Final remarks

This paper aimed to understand the multiple strategies frontline workers develop to deal with violence in vulnerable territories, characterized by high levels of violence, inequality, social and economic vulnerabilities. We undertook an analysis of the micro-dynamics within which frontline workers operate to understand the broader process of how the State deals with violence. The results suggest that the repertory of reactions developed by frontline workers is much more complex than suggested by previous studies (Diamond, 1997; Koritsas et al., 2010; McMahon et al., 2014; Robson et al., 2014; Zang, 2017). We found that, besides passivity in the face of violence, frontline workers also take positive actions and use their agency in order to affirmatively deal with violence.

This study has made four contributions to the existing literature about the subject. First, it expanded the understanding of how violence expresses itself in different ways in the territories in which policies are implemented. Therefore, even admitting that violence is a widespread and important phenomenon, we could see how it is expressed differently in different situations and affect the State and its policies in variable ways. Second, we reframed the relationship between violence, policy implementation, and frontline workers, restoring to the latter an active role in dealing with violence in multiple ways (Sundström, 2016). Even if studies about specific professions have been analyzing the consequences of violence in work (McMahon et al., 2014; Robson et al., 2014; Zang, 2017), the literature has reserved a passive or victim role for these workers. As we show here, violence can help in explaining the practices developed by these workers. Therefore, it can be a variable to understand agency, discretion, behavior, divergence, and coping. The third contribution, finally, is about the types of reactions mapped in the paper. Here, we were able to propose a typology of reactions and mechanisms through which frontline workers deal with these variables, which can be tested and applied in future studies. We recognize that this diversity is important for designing more adherent strategies to improve policy performance in contexts of violence, so common in countries and territories with high levels of inequality and vulnerability. The fourth contribution refers to identifying mechanisms that allow us to understand why one type of reaction or another prevails. Among other possible causes, we show how the territorial blindness or inserted character of policies and professionals can make the difference in restricting or expanding frontline workers' ability to deal affirmatively with violence, instead of simply submitting and being victims of it.

The main limitation of this study is the capacity to generalize the findings. As it was a non-representative sample of bureaucrats and, at the same time, the original research did not aim at understanding violence deeply, we cannot suggest that these findings are representative. Simultaneously, even this limitation is relevant to show how violence is central to these workers and how it appears recurrently and comprehensively. Using an extreme case, the findings may be of interest to similar contexts of vulnerable territories both in developing and developed countries.

Considering these limitations, future studies should test this typology in different contexts and with different types of workers. It would be interesting to see if the same can be found in other types of territories, particularly those contexts of high violence but less vulnerabilities or inequalities.